diff --git a/modules/core/perf/perf_dft.cpp b/modules/core/perf/perf_dft.cpp
index 05ae4bc50..a33bcf51e 100644
--- a/modules/core/perf/perf_dft.cpp
+++ b/modules/core/perf/perf_dft.cpp
@@ -22,5 +22,5 @@ PERF_TEST_P(Size_MatType, dft, TEST_MATS_DFT)
 
     TEST_CYCLE() dft(src, dst);
 
-    SANITY_CHECK(dst, 1e-5);
+    SANITY_CHECK(dst, 1e-5, ERROR_RELATIVE);
 }
diff --git a/modules/core/perf/perf_stat.cpp b/modules/core/perf/perf_stat.cpp
index 6b5f0ff52..25bad4f52 100644
--- a/modules/core/perf/perf_stat.cpp
+++ b/modules/core/perf/perf_stat.cpp
@@ -65,8 +65,8 @@ PERF_TEST_P(Size_MatType, meanStdDev, TYPICAL_MATS)
 
     TEST_CYCLE() meanStdDev(src, mean, dev);
 
-    SANITY_CHECK(mean, 1e-6);
-    SANITY_CHECK(dev, 1e-6);
+    SANITY_CHECK(mean, 1e-5, ERROR_RELATIVE);
+    SANITY_CHECK(dev, 1e-5, ERROR_RELATIVE);
 }
 
 PERF_TEST_P(Size_MatType, meanStdDev_mask, TYPICAL_MATS)
diff --git a/modules/core/src/arithm.cpp b/modules/core/src/arithm.cpp
index 17bd34cba..ecc2ca064 100644
--- a/modules/core/src/arithm.cpp
+++ b/modules/core/src/arithm.cpp
@@ -458,9 +458,12 @@ static void add8u( const uchar* src1, size_t step1,
                    const uchar* src2, size_t step2,
                    uchar* dst, size_t step, Size sz, void* )
 {
-    IF_IPP(fixSteps(sz, sizeof(dst[0]), step1, step2, step);
-           ippiAdd_8u_C1RSfs(src1, (int)step1, src2, (int)step2, dst, (int)step, (IppiSize&)sz, 0),
-           (vBinOp<uchar, OpAdd<uchar>, IF_SIMD(VAdd<uchar>)>(src1, step1, src2, step2, dst, step, sz)));
+#if (ARITHM_USE_IPP == 1)
+    fixSteps(sz, sizeof(dst[0]), step1, step2, step);
+    if (0 <= ippiAdd_8u_C1RSfs(src1, (int)step1, src2, (int)step2, dst, (int)step, (IppiSize&)sz, 0))
+        return;
+#endif
+    (vBinOp<uchar, OpAdd<uchar>, IF_SIMD(VAdd<uchar>)>(src1, step1, src2, step2, dst, step, sz));
 }
 
 static void add8s( const schar* src1, size_t step1,
@@ -474,18 +477,24 @@ static void add16u( const ushort* src1, size_t step1,
                     const ushort* src2, size_t step2,
                     ushort* dst, size_t step, Size sz, void* )
 {
-    IF_IPP(fixSteps(sz, sizeof(dst[0]), step1, step2, step);
-           ippiAdd_16u_C1RSfs(src1, (int)step1, src2, (int)step2, dst, (int)step, (IppiSize&)sz, 0),
-           (vBinOp<ushort, OpAdd<ushort>, IF_SIMD(VAdd<ushort>)>(src1, step1, src2, step2, dst, step, sz)));
+#if (ARITHM_USE_IPP == 1)
+    fixSteps(sz, sizeof(dst[0]), step1, step2, step);
+    if (0 <= ippiAdd_16u_C1RSfs(src1, (int)step1, src2, (int)step2, dst, (int)step, (IppiSize&)sz, 0))
+        return;
+#endif
+    (vBinOp<ushort, OpAdd<ushort>, IF_SIMD(VAdd<ushort>)>(src1, step1, src2, step2, dst, step, sz));
 }
 
 static void add16s( const short* src1, size_t step1,
                     const short* src2, size_t step2,
                     short* dst, size_t step, Size sz, void* )
 {
-    IF_IPP(fixSteps(sz, sizeof(dst[0]), step1, step2, step);
-           ippiAdd_16s_C1RSfs(src1, (int)step1, src2, (int)step2, dst, (int)step, (IppiSize&)sz, 0),
-           (vBinOp<short, OpAdd<short>, IF_SIMD(VAdd<short>)>(src1, step1, src2, step2, dst, step, sz)));
+#if (ARITHM_USE_IPP == 1)
+    fixSteps(sz, sizeof(dst[0]), step1, step2, step);
+    if (0 <= ippiAdd_16s_C1RSfs(src1, (int)step1, src2, (int)step2, dst, (int)step, (IppiSize&)sz, 0))
+        return;
+#endif
+    (vBinOp<short, OpAdd<short>, IF_SIMD(VAdd<short>)>(src1, step1, src2, step2, dst, step, sz));
 }
 
 static void add32s( const int* src1, size_t step1,
@@ -499,9 +508,12 @@ static void add32f( const float* src1, size_t step1,
                     const float* src2, size_t step2,
                     float* dst, size_t step, Size sz, void* )
 {
-    IF_IPP(fixSteps(sz, sizeof(dst[0]), step1, step2, step);
-           ippiAdd_32f_C1R(src1, (int)step1, src2, (int)step2, dst, (int)step, (IppiSize&)sz),
-           (vBinOp32<float, OpAdd<float>, IF_SIMD(VAdd<float>)>(src1, step1, src2, step2, dst, step, sz)));
+#if (ARITHM_USE_IPP == 1)
+    fixSteps(sz, sizeof(dst[0]), step1, step2, step);
+    if (0 <= ippiAdd_32f_C1R(src1, (int)step1, src2, (int)step2, dst, (int)step, (IppiSize&)sz))
+        return;
+#endif
+    (vBinOp32<float, OpAdd<float>, IF_SIMD(VAdd<float>)>(src1, step1, src2, step2, dst, step, sz));
 }
 
 static void add64f( const double* src1, size_t step1,
@@ -515,9 +527,12 @@ static void sub8u( const uchar* src1, size_t step1,
                    const uchar* src2, size_t step2,
                    uchar* dst, size_t step, Size sz, void* )
 {
-    IF_IPP(fixSteps(sz, sizeof(dst[0]), step1, step2, step);
-           ippiSub_8u_C1RSfs(src2, (int)step2, src1, (int)step1, dst, (int)step, (IppiSize&)sz, 0),
-           (vBinOp<uchar, OpSub<uchar>, IF_SIMD(VSub<uchar>)>(src1, step1, src2, step2, dst, step, sz)));
+#if (ARITHM_USE_IPP == 1)
+    fixSteps(sz, sizeof(dst[0]), step1, step2, step);
+    if (0 <= ippiSub_8u_C1RSfs(src2, (int)step2, src1, (int)step1, dst, (int)step, (IppiSize&)sz, 0))
+        return;
+#endif
+    (vBinOp<uchar, OpSub<uchar>, IF_SIMD(VSub<uchar>)>(src1, step1, src2, step2, dst, step, sz));
 }
 
 static void sub8s( const schar* src1, size_t step1,
@@ -531,18 +546,24 @@ static void sub16u( const ushort* src1, size_t step1,
                     const ushort* src2, size_t step2,
                     ushort* dst, size_t step, Size sz, void* )
 {
-    IF_IPP(fixSteps(sz, sizeof(dst[0]), step1, step2, step);
-           ippiSub_16u_C1RSfs(src2, (int)step2, src1, (int)step1, dst, (int)step, (IppiSize&)sz, 0),
-           (vBinOp<ushort, OpSub<ushort>, IF_SIMD(VSub<ushort>)>(src1, step1, src2, step2, dst, step, sz)));
+#if (ARITHM_USE_IPP == 1)
+    fixSteps(sz, sizeof(dst[0]), step1, step2, step);
+    if (0 <= ippiSub_16u_C1RSfs(src2, (int)step2, src1, (int)step1, dst, (int)step, (IppiSize&)sz, 0))
+        return;
+#endif
+    (vBinOp<ushort, OpSub<ushort>, IF_SIMD(VSub<ushort>)>(src1, step1, src2, step2, dst, step, sz));
 }
 
 static void sub16s( const short* src1, size_t step1,
                     const short* src2, size_t step2,
                     short* dst, size_t step, Size sz, void* )
 {
-    IF_IPP(fixSteps(sz, sizeof(dst[0]), step1, step2, step);
-           ippiSub_16s_C1RSfs(src2, (int)step2, src1, (int)step1, dst, (int)step, (IppiSize&)sz, 0),
-           (vBinOp<short, OpSub<short>, IF_SIMD(VSub<short>)>(src1, step1, src2, step2, dst, step, sz)));
+#if (ARITHM_USE_IPP == 1)
+    fixSteps(sz, sizeof(dst[0]), step1, step2, step);
+    if (0 <= ippiSub_16s_C1RSfs(src2, (int)step2, src1, (int)step1, dst, (int)step, (IppiSize&)sz, 0))
+        return;
+#endif
+    (vBinOp<short, OpSub<short>, IF_SIMD(VSub<short>)>(src1, step1, src2, step2, dst, step, sz));
 }
 
 static void sub32s( const int* src1, size_t step1,
@@ -556,9 +577,12 @@ static void sub32f( const float* src1, size_t step1,
                    const float* src2, size_t step2,
                    float* dst, size_t step, Size sz, void* )
 {
-    IF_IPP(fixSteps(sz, sizeof(dst[0]), step1, step2, step);
-           ippiSub_32f_C1R(src2, (int)step2, src1, (int)step1, dst, (int)step, (IppiSize&)sz),
-           (vBinOp32<float, OpSub<float>, IF_SIMD(VSub<float>)>(src1, step1, src2, step2, dst, step, sz)));
+#if (ARITHM_USE_IPP == 1)
+    fixSteps(sz, sizeof(dst[0]), step1, step2, step);
+    if (0 <= ippiSub_32f_C1R(src2, (int)step2, src1, (int)step1, dst, (int)step, (IppiSize&)sz))
+        return;
+#endif
+    (vBinOp32<float, OpSub<float>, IF_SIMD(VSub<float>)>(src1, step1, src2, step2, dst, step, sz));
 }
 
 static void sub64f( const double* src1, size_t step1,
@@ -576,26 +600,23 @@ static void max8u( const uchar* src1, size_t step1,
                    uchar* dst, size_t step, Size sz, void* )
 {
 #if (ARITHM_USE_IPP == 1)
-  {
     uchar* s1 = (uchar*)src1;
     uchar* s2 = (uchar*)src2;
     uchar* d  = dst;
     fixSteps(sz, sizeof(dst[0]), step1, step2, step);
-    for(int i = 0; i < sz.height; i++)
+    int i = 0;
+    for(; i < sz.height; i++)
     {
-      ippsMaxEvery_8u(s1, s2, d, sz.width);
-      s1 += step1;
-      s2 += step2;
-      d  += step;
+        if (0 > ippsMaxEvery_8u(s1, s2, d, sz.width))
+            break;
+        s1 += step1;
+        s2 += step2;
+        d  += step;
     }
-  }
-#else
-  vBinOp<uchar, OpMax<uchar>, IF_SIMD(VMax<uchar>)>(src1, step1, src2, step2, dst, step, sz);
+    if (i == sz.height)
+        return;
 #endif
-
-//    IF_IPP(fixSteps(sz, sizeof(dst[0]), step1, step2, step);
-//           ippiMaxEvery_8u_C1R(src1, (int)step1, src2, (int)step2, dst, (IppiSize&)sz),
-//           (vBinOp8<uchar, OpMax<uchar>, IF_SIMD(_VMax8u)>(src1, step1, src2, step2, dst, step, sz)));
+    vBinOp<uchar, OpMax<uchar>, IF_SIMD(VMax<uchar>)>(src1, step1, src2, step2, dst, step, sz);
 }
 
 static void max8s( const schar* src1, size_t step1,
@@ -610,26 +631,23 @@ static void max16u( const ushort* src1, size_t step1,
                     ushort* dst, size_t step, Size sz, void* )
 {
 #if (ARITHM_USE_IPP == 1)
-  {
     ushort* s1 = (ushort*)src1;
     ushort* s2 = (ushort*)src2;
     ushort* d  = dst;
     fixSteps(sz, sizeof(dst[0]), step1, step2, step);
-    for(int i = 0; i < sz.height; i++)
+    int i = 0;
+    for(; i < sz.height; i++)
     {
-      ippsMaxEvery_16u(s1, s2, d, sz.width);
-      s1 = (ushort*)((uchar*)s1 + step1);
-      s2 = (ushort*)((uchar*)s2 + step2);
-      d  = (ushort*)((uchar*)d + step);
+        if (0 > ippsMaxEvery_16u(s1, s2, d, sz.width))
+            break;
+        s1 = (ushort*)((uchar*)s1 + step1);
+        s2 = (ushort*)((uchar*)s2 + step2);
+        d  = (ushort*)((uchar*)d + step);
     }
-  }
-#else
-  vBinOp<ushort, OpMax<ushort>, IF_SIMD(VMax<ushort>)>(src1, step1, src2, step2, dst, step, sz);
+    if (i == sz.height)
+        return;
 #endif
-
-//    IF_IPP(fixSteps(sz, sizeof(dst[0]), step1, step2, step);
-//           ippiMaxEvery_16u_C1R(src1, (int)step1, src2, (int)step2, dst, (IppiSize&)sz),
-//           (vBinOp16<ushort, OpMax<ushort>, IF_SIMD(_VMax16u)>(src1, step1, src2, step2, dst, step, sz)));
+    vBinOp<ushort, OpMax<ushort>, IF_SIMD(VMax<ushort>)>(src1, step1, src2, step2, dst, step, sz);
 }
 
 static void max16s( const short* src1, size_t step1,
@@ -651,25 +669,23 @@ static void max32f( const float* src1, size_t step1,
                     float* dst, size_t step, Size sz, void* )
 {
 #if (ARITHM_USE_IPP == 1)
-  {
     float* s1 = (float*)src1;
     float* s2 = (float*)src2;
     float* d  = dst;
     fixSteps(sz, sizeof(dst[0]), step1, step2, step);
-    for(int i = 0; i < sz.height; i++)
+    int i = 0;
+    for(; i < sz.height; i++)
     {
-      ippsMaxEvery_32f(s1, s2, d, sz.width);
-      s1 = (float*)((uchar*)s1 + step1);
-      s2 = (float*)((uchar*)s2 + step2);
-      d  = (float*)((uchar*)d + step);
+        if (0 > ippsMaxEvery_32f(s1, s2, d, sz.width))
+            break;
+        s1 = (float*)((uchar*)s1 + step1);
+        s2 = (float*)((uchar*)s2 + step2);
+        d  = (float*)((uchar*)d + step);
     }
-  }
-#else
-  vBinOp32<float, OpMax<float>, IF_SIMD(VMax<float>)>(src1, step1, src2, step2, dst, step, sz);
+    if (i == sz.height)
+        return;
 #endif
-//    IF_IPP(fixSteps(sz, sizeof(dst[0]), step1, step2, step);
-//           ippiMaxEvery_32f_C1R(src1, (int)step1, src2, (int)step2, dst, (IppiSize&)sz),
-//           (vBinOp32f<OpMax<float>, IF_SIMD(_VMax32f)>(src1, step1, src2, step2, dst, step, sz)));
+    vBinOp32<float, OpMax<float>, IF_SIMD(VMax<float>)>(src1, step1, src2, step2, dst, step, sz);
 }
 
 static void max64f( const double* src1, size_t step1,
@@ -684,26 +700,23 @@ static void min8u( const uchar* src1, size_t step1,
                    uchar* dst, size_t step, Size sz, void* )
 {
 #if (ARITHM_USE_IPP == 1)
-  {
     uchar* s1 = (uchar*)src1;
     uchar* s2 = (uchar*)src2;
     uchar* d  = dst;
     fixSteps(sz, sizeof(dst[0]), step1, step2, step);
-    for(int i = 0; i < sz.height; i++)
+    int i = 0;
+    for(; i < sz.height; i++)
     {
-      ippsMinEvery_8u(s1, s2, d, sz.width);
-      s1 += step1;
-      s2 += step2;
-      d  += step;
+        if (0 > ippsMinEvery_8u(s1, s2, d, sz.width))
+            break;
+        s1 += step1;
+        s2 += step2;
+        d  += step;
     }
-  }
-#else
-  vBinOp<uchar, OpMin<uchar>, IF_SIMD(VMin<uchar>)>(src1, step1, src2, step2, dst, step, sz);
+    if (i == sz.height)
+        return;
 #endif
-
-//    IF_IPP(fixSteps(sz, sizeof(dst[0]), step1, step2, step);
-//           ippiMinEvery_8u_C1R(src1, (int)step1, src2, (int)step2, dst, (IppiSize&)sz),
-//           (vBinOp8<uchar, OpMin<uchar>, IF_SIMD(_VMin8u)>(src1, step1, src2, step2, dst, step, sz)));
+    vBinOp<uchar, OpMin<uchar>, IF_SIMD(VMin<uchar>)>(src1, step1, src2, step2, dst, step, sz);
 }
 
 static void min8s( const schar* src1, size_t step1,
@@ -718,26 +731,23 @@ static void min16u( const ushort* src1, size_t step1,
                     ushort* dst, size_t step, Size sz, void* )
 {
 #if (ARITHM_USE_IPP == 1)
-  {
     ushort* s1 = (ushort*)src1;
     ushort* s2 = (ushort*)src2;
     ushort* d  = dst;
     fixSteps(sz, sizeof(dst[0]), step1, step2, step);
-    for(int i = 0; i < sz.height; i++)
+    int i = 0;
+    for(; i < sz.height; i++)
     {
-      ippsMinEvery_16u(s1, s2, d, sz.width);
-      s1 = (ushort*)((uchar*)s1 + step1);
-      s2 = (ushort*)((uchar*)s2 + step2);
-      d  = (ushort*)((uchar*)d + step);
+        if (0 > ippsMinEvery_16u(s1, s2, d, sz.width))
+            break;
+        s1 = (ushort*)((uchar*)s1 + step1);
+        s2 = (ushort*)((uchar*)s2 + step2);
+        d  = (ushort*)((uchar*)d + step);
     }
-  }
-#else
-  vBinOp<ushort, OpMin<ushort>, IF_SIMD(VMin<ushort>)>(src1, step1, src2, step2, dst, step, sz);
+    if (i == sz.height)
+        return;
 #endif
-
-//    IF_IPP(fixSteps(sz, sizeof(dst[0]), step1, step2, step);
-//           ippiMinEvery_16u_C1R(src1, (int)step1, src2, (int)step2, dst, (IppiSize&)sz),
-//           (vBinOp16<ushort, OpMin<ushort>, IF_SIMD(_VMin16u)>(src1, step1, src2, step2, dst, step, sz)));
+    vBinOp<ushort, OpMin<ushort>, IF_SIMD(VMin<ushort>)>(src1, step1, src2, step2, dst, step, sz);
 }
 
 static void min16s( const short* src1, size_t step1,
@@ -759,25 +769,23 @@ static void min32f( const float* src1, size_t step1,
                     float* dst, size_t step, Size sz, void* )
 {
 #if (ARITHM_USE_IPP == 1)
-  {
     float* s1 = (float*)src1;
     float* s2 = (float*)src2;
     float* d  = dst;
     fixSteps(sz, sizeof(dst[0]), step1, step2, step);
-    for(int i = 0; i < sz.height; i++)
+    int i = 0;
+    for(; i < sz.height; i++)
     {
-      ippsMinEvery_32f(s1, s2, d, sz.width);
-      s1 = (float*)((uchar*)s1 + step1);
-      s2 = (float*)((uchar*)s2 + step2);
-      d  = (float*)((uchar*)d + step);
+        if (0 > ippsMinEvery_32f(s1, s2, d, sz.width))
+            break;
+        s1 = (float*)((uchar*)s1 + step1);
+        s2 = (float*)((uchar*)s2 + step2);
+        d  = (float*)((uchar*)d + step);
     }
-  }
-#else
-  vBinOp32<float, OpMin<float>, IF_SIMD(VMin<float>)>(src1, step1, src2, step2, dst, step, sz);
+    if (i == sz.height)
+        return;
 #endif
-//    IF_IPP(fixSteps(sz, sizeof(dst[0]), step1, step2, step);
-//           ippiMinEvery_32f_C1R(src1, (int)step1, src2, (int)step2, dst, (IppiSize&)sz),
-//           (vBinOp32f<OpMin<float>, IF_SIMD(_VMin32f)>(src1, step1, src2, step2, dst, step, sz)));
+    vBinOp32<float, OpMin<float>, IF_SIMD(VMin<float>)>(src1, step1, src2, step2, dst, step, sz);
 }
 
 static void min64f( const double* src1, size_t step1,
@@ -791,9 +799,12 @@ static void absdiff8u( const uchar* src1, size_t step1,
                        const uchar* src2, size_t step2,
                        uchar* dst, size_t step, Size sz, void* )
 {
-    IF_IPP(fixSteps(sz, sizeof(dst[0]), step1, step2, step);
-           ippiAbsDiff_8u_C1R(src1, (int)step1, src2, (int)step2, dst, (int)step, (IppiSize&)sz),
-           (vBinOp<uchar, OpAbsDiff<uchar>, IF_SIMD(VAbsDiff<uchar>)>(src1, step1, src2, step2, dst, step, sz)));
+#if (ARITHM_USE_IPP == 1)
+    fixSteps(sz, sizeof(dst[0]), step1, step2, step);
+    if (0 <= ippiAbsDiff_8u_C1R(src1, (int)step1, src2, (int)step2, dst, (int)step, (IppiSize&)sz))
+        return;
+#endif
+    (vBinOp<uchar, OpAbsDiff<uchar>, IF_SIMD(VAbsDiff<uchar>)>(src1, step1, src2, step2, dst, step, sz));
 }
 
 static void absdiff8s( const schar* src1, size_t step1,
@@ -807,9 +818,12 @@ static void absdiff16u( const ushort* src1, size_t step1,
                         const ushort* src2, size_t step2,
                         ushort* dst, size_t step, Size sz, void* )
 {
-    IF_IPP(fixSteps(sz, sizeof(dst[0]), step1, step2, step);
-           ippiAbsDiff_16u_C1R(src1, (int)step1, src2, (int)step2, dst, (int)step, (IppiSize&)sz),
-           (vBinOp<ushort, OpAbsDiff<ushort>, IF_SIMD(VAbsDiff<ushort>)>(src1, step1, src2, step2, dst, step, sz)));
+#if (ARITHM_USE_IPP == 1)
+    fixSteps(sz, sizeof(dst[0]), step1, step2, step);
+    if (0 <= ippiAbsDiff_16u_C1R(src1, (int)step1, src2, (int)step2, dst, (int)step, (IppiSize&)sz))
+        return;
+#endif
+    (vBinOp<ushort, OpAbsDiff<ushort>, IF_SIMD(VAbsDiff<ushort>)>(src1, step1, src2, step2, dst, step, sz));
 }
 
 static void absdiff16s( const short* src1, size_t step1,
@@ -830,9 +844,12 @@ static void absdiff32f( const float* src1, size_t step1,
                         const float* src2, size_t step2,
                         float* dst, size_t step, Size sz, void* )
 {
-    IF_IPP(fixSteps(sz, sizeof(dst[0]), step1, step2, step);
-           ippiAbsDiff_32f_C1R(src1, (int)step1, src2, (int)step2, dst, (int)step, (IppiSize&)sz),
-           (vBinOp32<float, OpAbsDiff<float>, IF_SIMD(VAbsDiff<float>)>(src1, step1, src2, step2, dst, step, sz)));
+#if (ARITHM_USE_IPP == 1)
+    fixSteps(sz, sizeof(dst[0]), step1, step2, step);
+    if (0 <= ippiAbsDiff_32f_C1R(src1, (int)step1, src2, (int)step2, dst, (int)step, (IppiSize&)sz))
+        return;
+#endif
+    (vBinOp32<float, OpAbsDiff<float>, IF_SIMD(VAbsDiff<float>)>(src1, step1, src2, step2, dst, step, sz));
 }
 
 static void absdiff64f( const double* src1, size_t step1,
@@ -847,36 +864,48 @@ static void and8u( const uchar* src1, size_t step1,
                    const uchar* src2, size_t step2,
                    uchar* dst, size_t step, Size sz, void* )
 {
-    IF_IPP(fixSteps(sz, sizeof(dst[0]), step1, step2, step);
-           ippiAnd_8u_C1R(src1, (int)step1, src2, (int)step2, dst, (int)step, (IppiSize&)sz),
-           (vBinOp<uchar, OpAnd<uchar>, IF_SIMD(VAnd<uchar>)>(src1, step1, src2, step2, dst, step, sz)));
+#if (ARITHM_USE_IPP == 1)
+    fixSteps(sz, sizeof(dst[0]), step1, step2, step);
+    if (0 <= ippiAnd_8u_C1R(src1, (int)step1, src2, (int)step2, dst, (int)step, (IppiSize&)sz))
+        return;
+#endif
+    (vBinOp<uchar, OpAnd<uchar>, IF_SIMD(VAnd<uchar>)>(src1, step1, src2, step2, dst, step, sz));
 }
 
 static void or8u( const uchar* src1, size_t step1,
                   const uchar* src2, size_t step2,
                   uchar* dst, size_t step, Size sz, void* )
 {
-    IF_IPP(fixSteps(sz, sizeof(dst[0]), step1, step2, step);
-           ippiOr_8u_C1R(src1, (int)step1, src2, (int)step2, dst, (int)step, (IppiSize&)sz),
-           (vBinOp<uchar, OpOr<uchar>, IF_SIMD(VOr<uchar>)>(src1, step1, src2, step2, dst, step, sz)));
+#if (ARITHM_USE_IPP == 1)
+    fixSteps(sz, sizeof(dst[0]), step1, step2, step);
+    if (0 <= ippiOr_8u_C1R(src1, (int)step1, src2, (int)step2, dst, (int)step, (IppiSize&)sz))
+        return;
+#endif
+    (vBinOp<uchar, OpOr<uchar>, IF_SIMD(VOr<uchar>)>(src1, step1, src2, step2, dst, step, sz));
 }
 
 static void xor8u( const uchar* src1, size_t step1,
                    const uchar* src2, size_t step2,
                    uchar* dst, size_t step, Size sz, void* )
 {
-    IF_IPP(fixSteps(sz, sizeof(dst[0]), step1, step2, step);
-           ippiXor_8u_C1R(src1, (int)step1, src2, (int)step2, dst, (int)step, (IppiSize&)sz),
-           (vBinOp<uchar, OpXor<uchar>, IF_SIMD(VXor<uchar>)>(src1, step1, src2, step2, dst, step, sz)));
+#if (ARITHM_USE_IPP == 1)
+    fixSteps(sz, sizeof(dst[0]), step1, step2, step);
+    if (0 <= ippiXor_8u_C1R(src1, (int)step1, src2, (int)step2, dst, (int)step, (IppiSize&)sz))
+        return;
+#endif
+    (vBinOp<uchar, OpXor<uchar>, IF_SIMD(VXor<uchar>)>(src1, step1, src2, step2, dst, step, sz));
 }
 
 static void not8u( const uchar* src1, size_t step1,
                    const uchar* src2, size_t step2,
                    uchar* dst, size_t step, Size sz, void* )
 {
-    IF_IPP(fixSteps(sz, sizeof(dst[0]), step1, step2, step); (void *)src2;
-           ippiNot_8u_C1R(src1, (int)step1, dst, (int)step, (IppiSize&)sz),
-           (vBinOp<uchar, OpNot<uchar>, IF_SIMD(VNot<uchar>)>(src1, step1, src2, step2, dst, step, sz)));
+#if (ARITHM_USE_IPP == 1)
+    fixSteps(sz, sizeof(dst[0]), step1, step2, step); (void *)src2;
+    if (0 <= ippiNot_8u_C1R(src1, (int)step1, dst, (int)step, (IppiSize&)sz))
+        return;
+#endif
+    (vBinOp<uchar, OpNot<uchar>, IF_SIMD(VNot<uchar>)>(src1, step1, src2, step2, dst, step, sz));
 }
 
 /****************************************************************************************\
@@ -2357,7 +2386,7 @@ static void cmp8u(const uchar* src1, size_t step1, const uchar* src2, size_t ste
     if( op  >= 0 )
     {
         fixSteps(size, sizeof(dst[0]), step1, step2, step);
-        if( ippiCompare_8u_C1R(src1, (int)step1, src2, (int)step2, dst, (int)step, (IppiSize&)size, op) >= 0 )
+        if (0 <= ippiCompare_8u_C1R(src1, (int)step1, src2, (int)step2, dst, (int)step, (IppiSize&)size, op))
             return;
     }
 #endif
@@ -2440,7 +2469,7 @@ static void cmp16u(const ushort* src1, size_t step1, const ushort* src2, size_t
     if( op  >= 0 )
     {
         fixSteps(size, sizeof(dst[0]), step1, step2, step);
-        if( ippiCompare_16u_C1R(src1, (int)step1, src2, (int)step2, dst, (int)step, (IppiSize&)size, op) >= 0 )
+        if (0 <= ippiCompare_16u_C1R(src1, (int)step1, src2, (int)step2, dst, (int)step, (IppiSize&)size, op))
             return;
     }
 #endif
@@ -2455,7 +2484,7 @@ static void cmp16s(const short* src1, size_t step1, const short* src2, size_t st
     if( op  > 0 )
     {
         fixSteps(size, sizeof(dst[0]), step1, step2, step);
-        if( ippiCompare_16s_C1R(src1, (int)step1, src2, (int)step2, dst, (int)step, (IppiSize&)size, op) >= 0 )
+        if (0 <= ippiCompare_16s_C1R(src1, (int)step1, src2, (int)step2, dst, (int)step, (IppiSize&)size, op))
             return;
     }
 #endif
@@ -2561,7 +2590,7 @@ static void cmp32f(const float* src1, size_t step1, const float* src2, size_t st
     if( op  >= 0 )
     {
         fixSteps(size, sizeof(dst[0]), step1, step2, step);
-        if( ippiCompare_32f_C1R(src1, (int)step1, src2, (int)step2, dst, (int)step, (IppiSize&)size, op) >= 0 )
+        if (0 <= ippiCompare_32f_C1R(src1, (int)step1, src2, (int)step2, dst, (int)step, (IppiSize&)size, op))
             return;
     }
 #endif
diff --git a/modules/core/src/mathfuncs.cpp b/modules/core/src/mathfuncs.cpp
index 16df02caf..12ba4fa5b 100644
--- a/modules/core/src/mathfuncs.cpp
+++ b/modules/core/src/mathfuncs.cpp
@@ -812,8 +812,6 @@ typedef union
 }
 DBLINT;
 
-#ifndef HAVE_IPP
-
 #define EXPTAB_SCALE 6
 #define EXPTAB_MASK  ((1 << EXPTAB_SCALE) - 1)
 
@@ -1275,13 +1273,26 @@ static void Exp_64f( const double *_x, double *y, int n )
 #undef EXPTAB_MASK
 #undef EXPPOLY_32F_A0
 
-#else
+#ifdef HAVE_IPP
+static void Exp_32f_ipp(const float *x, float *y, int n)
+{
+    if (0 <= ippsExp_32f_A21(x, y, n))
+        return;
+    Exp_32f(x, y, n);
+}
 
-#define Exp_32f ippsExp_32f_A21
-#define Exp_64f ippsExp_64f_A50
+static void Exp_64f_ipp(const double *x, double *y, int n)
+{
+    if (0 <= ippsExp_64f_A50(x, y, n))
+        return;
+    Exp_64f(x, y, n);
+}
 
+#define Exp_32f Exp_32f_ipp
+#define Exp_64f Exp_64f_ipp
 #endif
 
+
 void exp( InputArray _src, OutputArray _dst )
 {
     int type = _src.type(), depth = _src.depth(), cn = _src.channels();
@@ -1302,9 +1313,9 @@ void exp( InputArray _src, OutputArray _dst )
     for( size_t i = 0; i < it.nplanes; i++, ++it )
     {
         if( depth == CV_32F )
-            Exp_32f( (const float*)ptrs[0], (float*)ptrs[1], len );
+            Exp_32f((const float*)ptrs[0], (float*)ptrs[1], len);
         else
-            Exp_64f( (const double*)ptrs[0], (double*)ptrs[1], len );
+            Exp_64f((const double*)ptrs[0], (double*)ptrs[1], len);
     }
 }
 
@@ -1313,8 +1324,6 @@ void exp( InputArray _src, OutputArray _dst )
 *                                          L O G                                         *
 \****************************************************************************************/
 
-#ifndef HAVE_IPP
-
 #define LOGTAB_SCALE    8
 #define LOGTAB_MASK         ((1 << LOGTAB_SCALE) - 1)
 #define LOGTAB_MASK2        ((1 << (20 - LOGTAB_SCALE)) - 1)
@@ -1922,11 +1931,23 @@ static void Log_64f( const double *x, double *y, int n )
     }
 }
 
-#else
+#ifdef HAVE_IPP
+static void Log_32f_ipp(const float *x, float *y, int n)
+{
+    if (0 <= ippsLn_32f_A21(x, y, n))
+        return;
+    Log_32f(x, y, n);
+}
 
-#define Log_32f ippsLn_32f_A21
-#define Log_64f ippsLn_64f_A50
+static void Log_64f_ipp(const double *x, double *y, int n)
+{
+    if (0 <= ippsLn_64f_A50(x, y, n))
+        return;
+    Log_64f(x, y, n);
+}
 
+#define Log_32f Log_32f_ipp
+#define Log_64f Log_64f_ipp
 #endif
 
 void log( InputArray _src, OutputArray _dst )
diff --git a/modules/core/src/matmul.cpp b/modules/core/src/matmul.cpp
index daad86a0f..8891bb05f 100644
--- a/modules/core/src/matmul.cpp
+++ b/modules/core/src/matmul.cpp
@@ -44,10 +44,6 @@
 #include "opencl_kernels.hpp"
 #include "opencv2/core/opencl/runtime/opencl_clamdblas.hpp"
 
-#ifdef HAVE_IPP
-#include "ippversion.h"
-#endif
-
 namespace cv
 {
 
@@ -2803,11 +2799,11 @@ static double dotProd_8u(const uchar* src1, const uchar* src2, int len)
 {
     double r = 0;
 #if ARITHM_USE_IPP
-    ippiDotProd_8u64f_C1R(src1, (int)(len*sizeof(src1[0])),
-                          src2, (int)(len*sizeof(src2[0])),
-                          ippiSize(len, 1), &r);
-    return r;
-#else
+    if (0 <= ippiDotProd_8u64f_C1R(src1, (int)(len*sizeof(src1[0])),
+                                   src2, (int)(len*sizeof(src2[0])),
+                                   ippiSize(len, 1), &r))
+        return r;
+#endif
     int i = 0;
 
 #if CV_SSE2
@@ -2853,7 +2849,6 @@ static double dotProd_8u(const uchar* src1, const uchar* src2, int len)
     }
 #endif
     return r + dotProd_(src1, src2, len - i);
-#endif
 }
 
 
@@ -2864,48 +2859,52 @@ static double dotProd_8s(const schar* src1, const schar* src2, int len)
 
 static double dotProd_16u(const ushort* src1, const ushort* src2, int len)
 {
+#if (ARITHM_USE_IPP == 1)
     double r = 0;
-    IF_IPP(ippiDotProd_16u64f_C1R(src1, (int)(len*sizeof(src1[0])),
-                                  src2, (int)(len*sizeof(src2[0])),
-                                  ippiSize(len, 1), &r),
-           r = dotProd_(src1, src2, len));
-    return r;
+    if (0 <= ippiDotProd_16u64f_C1R(src1, (int)(len*sizeof(src1[0])), src2, (int)(len*sizeof(src2[0])), ippiSize(len, 1), &r))
+        return r;
+#endif
+    return dotProd_(src1, src2, len);
 }
 
 static double dotProd_16s(const short* src1, const short* src2, int len)
 {
+#if (ARITHM_USE_IPP == 1)
     double r = 0;
-    IF_IPP(ippiDotProd_16s64f_C1R(src1, (int)(len*sizeof(src1[0])),
-                                  src2, (int)(len*sizeof(src2[0])),
-                                  ippiSize(len, 1), &r),
-           r = dotProd_(src1, src2, len));
-    return r;
+    if (0 <= ippiDotProd_16s64f_C1R(src1, (int)(len*sizeof(src1[0])), src2, (int)(len*sizeof(src2[0])), ippiSize(len, 1), &r))
+        return r;
+#endif
+    return dotProd_(src1, src2, len);
 }
 
 static double dotProd_32s(const int* src1, const int* src2, int len)
 {
+#if (ARITHM_USE_IPP == 1)
     double r = 0;
-    IF_IPP(ippiDotProd_32s64f_C1R(src1, (int)(len*sizeof(src1[0])),
-                                  src2, (int)(len*sizeof(src2[0])),
-                                  ippiSize(len, 1), &r),
-           r = dotProd_(src1, src2, len));
-    return r;
+    if (0 <= ippiDotProd_32s64f_C1R(src1, (int)(len*sizeof(src1[0])), src2, (int)(len*sizeof(src2[0])), ippiSize(len, 1), &r))
+        return r;
+#endif
+    return dotProd_(src1, src2, len);
 }
 
 static double dotProd_32f(const float* src1, const float* src2, int len)
 {
+#if (ARITHM_USE_IPP == 1)
     double r = 0;
-    IF_IPP(ippsDotProd_32f64f(src1, src2, len, &r),
-           r = dotProd_(src1, src2, len));
-    return r;
+    if (0 <= ippsDotProd_32f64f(src1, src2, len, &r))
+        return r;
+#endif
+    return dotProd_(src1, src2, len);
 }
 
 static double dotProd_64f(const double* src1, const double* src2, int len)
 {
+#if (ARITHM_USE_IPP == 1)
     double r = 0;
-    IF_IPP(ippsDotProd_64f(src1, src2, len, &r),
-           r = dotProd_(src1, src2, len));
-    return r;
+    if (0 <= ippsDotProd_64f(src1, src2, len, &r))
+        return r;
+#endif
+    return dotProd_(src1, src2, len);
 }
 
 
diff --git a/modules/core/src/precomp.hpp b/modules/core/src/precomp.hpp
index ff5943bc6..1c498cd01 100644
--- a/modules/core/src/precomp.hpp
+++ b/modules/core/src/precomp.hpp
@@ -199,10 +199,8 @@ enum { BLOCK_SIZE = 1024 };
 
 #if defined HAVE_IPP && (IPP_VERSION_MAJOR >= 7)
 #define ARITHM_USE_IPP 1
-#define IF_IPP(then_call, else_call) then_call
 #else
 #define ARITHM_USE_IPP 0
-#define IF_IPP(then_call, else_call) else_call
 #endif
 
 inline bool checkScalar(const Mat& sc, int atype, int sckind, int akind)
diff --git a/modules/core/src/stat.cpp b/modules/core/src/stat.cpp
index a4605d160..b7e5d3422 100644
--- a/modules/core/src/stat.cpp
+++ b/modules/core/src/stat.cpp
@@ -43,7 +43,6 @@
 #include "precomp.hpp"
 #include "opencl_kernels.hpp"
 #include <climits>
-#include <limits>
 
 namespace cv
 {
@@ -972,7 +971,9 @@ void cv::meanStdDev( InputArray _src, OutputArray _mean, OutputArray _sdv, Input
             ippiMeanStdDevFuncC1 ippFuncC1 =
             type == CV_8UC1 ? (ippiMeanStdDevFuncC1)ippiMean_StdDev_8u_C1R :
             type == CV_16UC1 ? (ippiMeanStdDevFuncC1)ippiMean_StdDev_16u_C1R :
-            //type == CV_32FC1 ? (ippiMeanStdDevFuncC1)ippiMean_StdDev_32f_C1R ://Aug 2013: bug in IPP 7.1, 8.0
+#if (IPP_VERSION_MAJOR >= 8) && (IPP_VERSION_MINOR >= 1)
+            type == CV_32FC1 ? (ippiMeanStdDevFuncC1)ippiMean_StdDev_32f_C1R ://Aug 2013: bug in IPP 7.1, 8.0
+#endif
             0;
             if( ippFuncC1 )
             {
@@ -2111,8 +2112,10 @@ double cv::norm( InputArray _src, int normType, InputArray _mask )
                 type == CV_16UC3 ? (ippiNormFuncNoHint)ippiNorm_Inf_16u_C3R :
                 type == CV_16UC4 ? (ippiNormFuncNoHint)ippiNorm_Inf_16u_C4R :
                 type == CV_16SC1 ? (ippiNormFuncNoHint)ippiNorm_Inf_16s_C1R :
-                //type == CV_16SC3 ? (ippiNormFunc)ippiNorm_Inf_16s_C3R : //Aug 2013: problem in IPP 7.1, 8.0 : -32768
-                //type == CV_16SC4 ? (ippiNormFunc)ippiNorm_Inf_16s_C4R : //Aug 2013: problem in IPP 7.1, 8.0 : -32768
+#if (IPP_VERSION_MAJOR >= 8) && (IPP_VERSION_MINOR >= 1)
+                type == CV_16SC3 ? (ippiNormFuncNoHint)ippiNorm_Inf_16s_C3R : //Aug 2013: problem in IPP 7.1, 8.0 : -32768
+                type == CV_16SC4 ? (ippiNormFuncNoHint)ippiNorm_Inf_16s_C4R : //Aug 2013: problem in IPP 7.1, 8.0 : -32768
+#endif
                 type == CV_32FC1 ? (ippiNormFuncNoHint)ippiNorm_Inf_32f_C1R :
                 type == CV_32FC3 ? (ippiNormFuncNoHint)ippiNorm_Inf_32f_C3R :
                 type == CV_32FC4 ? (ippiNormFuncNoHint)ippiNorm_Inf_32f_C4R :
@@ -2541,8 +2544,10 @@ double cv::norm( InputArray _src1, InputArray _src2, int normType, InputArray _m
                 type == CV_16UC3 ? (ippiNormDiffFuncNoHint)ippiNormDiff_Inf_16u_C3R :
                 type == CV_16UC4 ? (ippiNormDiffFuncNoHint)ippiNormDiff_Inf_16u_C4R :
                 type == CV_16SC1 ? (ippiNormDiffFuncNoHint)ippiNormDiff_Inf_16s_C1R :
-                //type == CV_16SC3 ? (ippiNormDiffFunc)ippiNormDiff_Inf_16s_C3R : //Aug 2013: problem in IPP 7.1, 8.0 : -32768
-                //type == CV_16SC4 ? (ippiNormDiffFunc)ippiNormDiff_Inf_16s_C4R : //Aug 2013: problem in IPP 7.1, 8.0 : -32768
+#if (IPP_VERSION_MAJOR >= 8) && (IPP_VERSION_MINOR >= 1)
+                type == CV_16SC3 ? (ippiNormDiffFuncNoHint)ippiNormDiff_Inf_16s_C3R : //Aug 2013: problem in IPP 7.1, 8.0 : -32768
+                type == CV_16SC4 ? (ippiNormDiffFuncNoHint)ippiNormDiff_Inf_16s_C4R : //Aug 2013: problem in IPP 7.1, 8.0 : -32768
+#endif
                 type == CV_32FC1 ? (ippiNormDiffFuncNoHint)ippiNormDiff_Inf_32f_C1R :
                 type == CV_32FC3 ? (ippiNormDiffFuncNoHint)ippiNormDiff_Inf_32f_C3R :
                 type == CV_32FC4 ? (ippiNormDiffFuncNoHint)ippiNormDiff_Inf_32f_C4R :
diff --git a/modules/core/src/system.cpp b/modules/core/src/system.cpp
index 5b8a69649..4b3efce4a 100644
--- a/modules/core/src/system.cpp
+++ b/modules/core/src/system.cpp
@@ -397,17 +397,17 @@ int64 getCPUTickCount(void)
 
 #else
 
-#ifdef HAVE_IPP
-int64 getCPUTickCount(void)
-{
-    return ippGetCpuClocks();
-}
-#else
+//#ifdef HAVE_IPP
+//int64 getCPUTickCount(void)
+//{
+//    return ippGetCpuClocks();
+//}
+//#else
 int64 getCPUTickCount(void)
 {
     return getTickCount();
 }
-#endif
+//#endif
 
 #endif
 
diff --git a/modules/imgproc/doc/feature_detection.rst b/modules/imgproc/doc/feature_detection.rst
index de16c0751..2187b8fd8 100644
--- a/modules/imgproc/doc/feature_detection.rst
+++ b/modules/imgproc/doc/feature_detection.rst
@@ -1,4 +1,4 @@
-Feature Detection
+Feature Detection
 =================
 
 .. highlight:: cpp
@@ -15,9 +15,9 @@ Finds edges in an image using the [Canny86]_ algorithm.
 
 .. ocv:cfunction:: void cvCanny( const CvArr* image, CvArr* edges, double threshold1, double threshold2, int aperture_size=3 )
 
-    :param image: single-channel 8-bit input image.
+    :param image: 8-bit input image.
 
-    :param edges: output edge map; it has the same size and type as  ``image`` .
+    :param edges: output edge map; single channels 8-bit image, which has the same size as  ``image`` .
 
     :param threshold1: first threshold for the hysteresis procedure.
 
diff --git a/modules/imgproc/perf/perf_bilateral.cpp b/modules/imgproc/perf/perf_bilateral.cpp
index af565f8a5..4ed0c4db6 100644
--- a/modules/imgproc/perf/perf_bilateral.cpp
+++ b/modules/imgproc/perf/perf_bilateral.cpp
@@ -34,5 +34,5 @@ PERF_TEST_P( TestBilateralFilter, BilateralFilter,
 
     TEST_CYCLE() bilateralFilter(src, dst, d, sigmaColor, sigmaSpace, BORDER_DEFAULT);
 
-    SANITY_CHECK(dst);
+    SANITY_CHECK(dst, .01, ERROR_RELATIVE);
 }
diff --git a/modules/imgproc/src/canny.cpp b/modules/imgproc/src/canny.cpp
index fbc92dde2..65e817306 100644
--- a/modules/imgproc/src/canny.cpp
+++ b/modules/imgproc/src/canny.cpp
@@ -42,13 +42,13 @@
 #include "precomp.hpp"
 #include "opencl_kernels.hpp"
 
-/*
+
 #if defined (HAVE_IPP) && (IPP_VERSION_MAJOR >= 7)
 #define USE_IPP_CANNY 1
 #else
 #undef USE_IPP_CANNY
 #endif
-*/
+
 
 namespace cv
 {
@@ -81,8 +81,8 @@ static bool ippCanny(const Mat& _src, Mat& _dst, float low,  float high)
         return false;
 
     if( ippiCanny_16s8u_C1R(_dx.ptr<short>(), (int)_dx.step,
-                            _dy.ptr<short>(), (int)_dy.step,
-                            _dst.data, (int)_dst.step, roi, low, high, buffer) < 0 )
+                               _dy.ptr<short>(), (int)_dy.step,
+                              _dst.data, (int)_dst.step, roi, low, high, buffer) < 0 )
         return false;
     return true;
 }
@@ -286,7 +286,7 @@ void cv::Canny( InputArray _src, OutputArray _dst,
 #endif
 
 #ifdef USE_IPP_CANNY
-    if( aperture_size == 3 && !L2gradient &&
+    if( aperture_size == 3 && !L2gradient && 1 == cn &&
         ippCanny(src, dst, (float)low_thresh, (float)high_thresh) )
         return;
 #endif
diff --git a/modules/imgproc/src/color.cpp b/modules/imgproc/src/color.cpp
index 3822ab3c1..30b04c238 100644
--- a/modules/imgproc/src/color.cpp
+++ b/modules/imgproc/src/color.cpp
@@ -252,6 +252,7 @@ bool CvtColorIPPLoopCopy(Mat& src, Mat& dst, const Cvt& cvt)
     }
     bool ok;
     parallel_for_(Range(0, source.rows), CvtColorIPPLoop_Invoker<Cvt>(source, dst, cvt, &ok), source.total()/(double)(1<<16) );
+    //ok = cvt(src.ptr<uchar>(0), (int)src.step[0], dst.ptr<uchar>(0), (int)dst.step[0], src.cols, src.rows);
     return ok;
 }
 
@@ -297,11 +298,13 @@ static ippiReorderFunc ippiSwapChannelsC3RTab[] =
     0, (ippiReorderFunc)ippiSwapChannels_32f_C3R, 0, 0
 };
 
+#if (IPP_VERSION_MAJOR >= 8) && (IPP_VERSION_MINOR >= 1)
 static ippiReorderFunc ippiSwapChannelsC4RTab[] =
 {
-    (ippiReorderFunc)ippiSwapChannels_8u_AC4R, 0, (ippiReorderFunc)ippiSwapChannels_16u_AC4R, 0,
-    0, (ippiReorderFunc)ippiSwapChannels_32f_AC4R, 0, 0
+    (ippiReorderFunc)ippiSwapChannels_8u_C4R, 0, (ippiReorderFunc)ippiSwapChannels_16u_C4R, 0,
+    0, (ippiReorderFunc)ippiSwapChannels_32f_C4R, 0, 0
 };
+#endif
 
 static ippiColor2GrayFunc ippiColor2GrayC3Tab[] =
 {
@@ -3251,11 +3254,13 @@ void cv::cvtColor( InputArray _src, OutputArray _dst, int code, int dcn )
                 if( CvtColorIPPLoopCopy(src, dst, IPPReorderFunctor(ippiSwapChannelsC3RTab[depth], 2, 1, 0)) )
                     return;
             }
+#if (IPP_VERSION_MAJOR >= 8) && (IPP_VERSION_MINOR >= 1)
             else if( code == CV_RGBA2BGRA )
             {
                 if( CvtColorIPPLoopCopy(src, dst, IPPReorderFunctor(ippiSwapChannelsC4RTab[depth], 2, 1, 0)) )
                     return;
             }
+#endif
 #endif
 
             if( depth == CV_8U )
@@ -3310,14 +3315,17 @@ void cv::cvtColor( InputArray _src, OutputArray _dst, int code, int dcn )
             CV_Assert( scn == 3 || scn == 4 );
             _dst.create(sz, CV_MAKETYPE(depth, 1));
             dst = _dst.getMat();
-/*
+/**/
 #if defined (HAVE_IPP) && (IPP_VERSION_MAJOR >= 7)
+/*
             if( code == CV_BGR2GRAY )
             {
                 if( CvtColorIPPLoop(src, dst, IPPColor2GrayFunctor(ippiColor2GrayC3Tab[depth])) )
                     return;
             }
-            else if( code == CV_RGB2GRAY )
+            else
+*/
+            if( code == CV_RGB2GRAY )
             {
                 if( CvtColorIPPLoop(src, dst, IPPGeneralFunctor(ippiRGB2GrayC3Tab[depth])) )
                     return;
@@ -3333,7 +3341,7 @@ void cv::cvtColor( InputArray _src, OutputArray _dst, int code, int dcn )
                     return;
             }
 #endif
-*/
+/**/
             bidx = code == CV_BGR2GRAY || code == CV_BGRA2GRAY ? 0 : 2;
 
             if( depth == CV_8U )
diff --git a/modules/imgproc/src/deriv.cpp b/modules/imgproc/src/deriv.cpp
index df2f371f5..0b19f22be 100644
--- a/modules/imgproc/src/deriv.cpp
+++ b/modules/imgproc/src/deriv.cpp
@@ -190,223 +190,231 @@ namespace cv
 
 static bool IPPDerivScharr(const Mat& src, Mat& dst, int ddepth, int dx, int dy, double scale)
 {
-   int bufSize = 0;
-   cv::AutoBuffer<char> buffer;
-   IppiSize roi = ippiSize(src.cols, src.rows);
+    int bufSize = 0;
+    cv::AutoBuffer<char> buffer;
+    IppiSize roi = ippiSize(src.cols, src.rows);
 
-   if( ddepth < 0 )
-     ddepth = src.depth();
+    if( ddepth < 0 )
+        ddepth = src.depth();
 
-   dst.create( src.size(), CV_MAKETYPE(ddepth, src.channels()) );
+    dst.create( src.size(), CV_MAKETYPE(ddepth, src.channels()) );
 
-   switch(src.type())
-   {
-      case CV_8U:
-         {
+    switch(src.type())
+    {
+    case CV_8U:
+        {
             if(scale != 1)
                 return false;
 
             switch(dst.type())
             {
-               case CV_16S:
-               {
-                  if((dx == 1) && (dy == 0))
-                  {
-                     ippiFilterScharrVertGetBufferSize_8u16s_C1R(roi,&bufSize);
-                     buffer.allocate(bufSize);
-
-                     ippiFilterScharrVertBorder_8u16s_C1R((const Ipp8u*)src.data, (int)src.step,
-                        (Ipp16s*)dst.data, (int)dst.step, roi, ippBorderRepl, 0, (Ipp8u*)(char*)buffer);
-
-                     return true;
-                  }
-
-                  if((dx == 0) && (dy == 1))
-                  {
-                     ippiFilterScharrHorizGetBufferSize_8u16s_C1R(roi,&bufSize);
-                     buffer.allocate(bufSize);
-
-                     ippiFilterScharrHorizBorder_8u16s_C1R((const Ipp8u*)src.data, (int)src.step,
-                        (Ipp16s*)dst.data, (int)dst.step, roi, ippBorderRepl, 0, (Ipp8u*)(char*)buffer);
-
-                     return true;
-                  }
-               }
-
-               default:
-                  return false;
+            case CV_16S:
+                {
+                    if ((dx == 1) && (dy == 0))
+                    {
+                        if (0 > ippiFilterScharrVertGetBufferSize_8u16s_C1R(roi,&bufSize))
+                            return false;
+                        buffer.allocate(bufSize);
+                        return (0 <= ippiFilterScharrVertBorder_8u16s_C1R((const Ipp8u*)src.data, (int)src.step,
+                                        (Ipp16s*)dst.data, (int)dst.step, roi, ippBorderRepl, 0, (Ipp8u*)(char*)buffer));
+                    }
+                    if ((dx == 0) && (dy == 1))
+                    {
+                        if (0 > ippiFilterScharrHorizGetBufferSize_8u16s_C1R(roi,&bufSize))
+                            return false;
+                        buffer.allocate(bufSize);
+                        return (0 <= ippiFilterScharrHorizBorder_8u16s_C1R((const Ipp8u*)src.data, (int)src.step,
+                                            (Ipp16s*)dst.data, (int)dst.step, roi, ippBorderRepl, 0, (Ipp8u*)(char*)buffer));
+                    }
+                    return false;
+                }
+            default:
+                return false;
             }
-         }
-
-      case CV_32F:
-         {
+        }
+    case CV_32F:
+        {
             switch(dst.type())
             {
-               case CV_32F:
-               if((dx == 1) && (dy == 0))
-               {
-                  ippiFilterScharrVertGetBufferSize_32f_C1R(ippiSize(src.cols, src.rows),&bufSize);
-                  buffer.allocate(bufSize);
+            case CV_32F:
+                {
+                    if ((dx == 1) && (dy == 0))
+                    {
+                        if (0 > ippiFilterScharrVertGetBufferSize_32f_C1R(ippiSize(src.cols, src.rows),&bufSize))
+                            return false;
+                        buffer.allocate(bufSize);
 
-                  ippiFilterScharrVertBorder_32f_C1R((const Ipp32f*)src.data, (int)src.step,
-                     (Ipp32f*)dst.data, (int)dst.step, ippiSize(src.cols, src.rows),
-                                            ippBorderRepl, 0, (Ipp8u*)(char*)buffer);
-                  if(scale != 1)
-                     /* IPP is fast, so MulC produce very little perf degradation */
-                     ippiMulC_32f_C1IR((Ipp32f)scale, (Ipp32f*)dst.data, (int)dst.step, ippiSize(dst.cols*dst.channels(), dst.rows));
+                        if (0 > ippiFilterScharrVertBorder_32f_C1R((const Ipp32f*)src.data, (int)src.step,
+                                        (Ipp32f*)dst.data, (int)dst.step, ippiSize(src.cols, src.rows),
+                                        ippBorderRepl, 0, (Ipp8u*)(char*)buffer))
+                        {
+                            return false;
+                        }
 
-                  return true;
-               }
+                        if (scale != 1)
+                            /* IPP is fast, so MulC produce very little perf degradation.*/
+                            //ippiMulC_32f_C1IR((Ipp32f)scale, (Ipp32f*)dst.data, (int)dst.step, ippiSize(dst.cols*dst.channels(), dst.rows));
+                            ippiMulC_32f_C1R((Ipp32f*)dst.data, (int)dst.step, (Ipp32f)scale, (Ipp32f*)dst.data, (int)dst.step, ippiSize(dst.cols*dst.channels(), dst.rows));
+                        return true;
+                    }
+                    if ((dx == 0) && (dy == 1))
+                    {
+                        if (0 > ippiFilterScharrHorizGetBufferSize_32f_C1R(ippiSize(src.cols, src.rows),&bufSize))
+                            return false;
+                        buffer.allocate(bufSize);
 
-               if((dx == 0) && (dy == 1))
-               {
-                  ippiFilterScharrHorizGetBufferSize_32f_C1R(ippiSize(src.cols, src.rows),&bufSize);
-                  buffer.allocate(bufSize);
+                        if (0 > ippiFilterScharrHorizBorder_32f_C1R((const Ipp32f*)src.data, (int)src.step,
+                                        (Ipp32f*)dst.data, (int)dst.step, ippiSize(src.cols, src.rows),
+                                        ippBorderRepl, 0, (Ipp8u*)(char*)buffer))
+                            return false;
 
-                  ippiFilterScharrHorizBorder_32f_C1R((const Ipp32f*)src.data, (int)src.step,
-                     (Ipp32f*)dst.data, (int)dst.step, ippiSize(src.cols, src.rows),
-                                            ippBorderRepl, 0, (Ipp8u*)(char*)buffer);
-                  if(scale != 1)
-                     ippiMulC_32f_C1IR((Ipp32f)scale, (Ipp32f *)dst.data, (int)dst.step, ippiSize(dst.cols*dst.channels(), dst.rows));
-
-                  return true;
-               }
-
-               default:
-                  return false;
+                        if (scale != 1)
+                            ippiMulC_32f_C1R((Ipp32f *)dst.data, (int)dst.step, (Ipp32f)scale, (Ipp32f *)dst.data, (int)dst.step, ippiSize(dst.cols*dst.channels(), dst.rows));
+                        return true;
+                    }
+                }
+            default:
+                return false;
             }
-         }
-
-      default:
-         return false;
-   }
+        }
+    default:
+        return false;
+    }
 }
 
 
 static bool IPPDeriv(const Mat& src, Mat& dst, int ddepth, int dx, int dy, int ksize, double scale)
 {
-   int bufSize = 0;
-   cv::AutoBuffer<char> buffer;
+    int bufSize = 0;
+    cv::AutoBuffer<char> buffer;
+    if (ksize == 3 || ksize == 5)
+    {
+        if ( ddepth < 0 )
+            ddepth = src.depth();
 
-   if(ksize == 3 || ksize == 5)
-   {
-      if( ddepth < 0 )
-          ddepth = src.depth();
+        if (src.type() == CV_8U && dst.type() == CV_16S && scale == 1)
+        {
+            if ((dx == 1) && (dy == 0))
+            {
+                if (0 > ippiFilterSobelNegVertGetBufferSize_8u16s_C1R(ippiSize(src.cols, src.rows), (IppiMaskSize)(ksize*10+ksize),&bufSize))
+                    return false;
+                buffer.allocate(bufSize);
 
-      if(src.type() == CV_8U && dst.type() == CV_16S && scale == 1)
-      {
-         if((dx == 1) && (dy == 0))
-         {
-            ippiFilterSobelNegVertGetBufferSize_8u16s_C1R(ippiSize(src.cols, src.rows), (IppiMaskSize)(ksize*10+ksize),&bufSize);
-            buffer.allocate(bufSize);
+                return (0 <= ippiFilterSobelNegVertBorder_8u16s_C1R((const Ipp8u*)src.data, (int)src.step,
+                                    (Ipp16s*)dst.data, (int)dst.step, ippiSize(src.cols, src.rows), (IppiMaskSize)(ksize*10+ksize),
+                                    ippBorderRepl, 0, (Ipp8u*)(char*)buffer));
+            }
 
-            ippiFilterSobelNegVertBorder_8u16s_C1R((const Ipp8u*)src.data, (int)src.step,
-               (Ipp16s*)dst.data, (int)dst.step, ippiSize(src.cols, src.rows), (IppiMaskSize)(ksize*10+ksize),
-                                      ippBorderRepl, 0, (Ipp8u*)(char*)buffer);
-            return true;
-         }
+            if ((dx == 0) && (dy == 1))
+            {
+                if (0 > ippiFilterSobelHorizGetBufferSize_8u16s_C1R(ippiSize(src.cols, src.rows), (IppiMaskSize)(ksize*10+ksize),&bufSize))
+                    return false;
+                buffer.allocate(bufSize);
 
-         if((dx == 0) && (dy == 1))
-         {
-            ippiFilterSobelHorizGetBufferSize_8u16s_C1R(ippiSize(src.cols, src.rows), (IppiMaskSize)(ksize*10+ksize),&bufSize);
-            buffer.allocate(bufSize);
+                return (0 <= ippiFilterSobelHorizBorder_8u16s_C1R((const Ipp8u*)src.data, (int)src.step,
+                                    (Ipp16s*)dst.data, (int)dst.step, ippiSize(src.cols, src.rows), (IppiMaskSize)(ksize*10+ksize),
+                                    ippBorderRepl, 0, (Ipp8u*)(char*)buffer));
+            }
 
-            ippiFilterSobelHorizBorder_8u16s_C1R((const Ipp8u*)src.data, (int)src.step,
-               (Ipp16s*)dst.data, (int)dst.step, ippiSize(src.cols, src.rows), (IppiMaskSize)(ksize*10+ksize),
-                                      ippBorderRepl, 0, (Ipp8u*)(char*)buffer);
+            if ((dx == 2) && (dy == 0))
+            {
+                if (0 > ippiFilterSobelVertSecondGetBufferSize_8u16s_C1R(ippiSize(src.cols, src.rows), (IppiMaskSize)(ksize*10+ksize),&bufSize))
+                    return false;
+                buffer.allocate(bufSize);
 
-            return true;
-         }
+                return (0 <= ippiFilterSobelVertSecondBorder_8u16s_C1R((const Ipp8u*)src.data, (int)src.step,
+                                    (Ipp16s*)dst.data, (int)dst.step, ippiSize(src.cols, src.rows), (IppiMaskSize)(ksize*10+ksize),
+                                    ippBorderRepl, 0, (Ipp8u*)(char*)buffer));
+            }
 
-         if((dx == 2) && (dy == 0))
-         {
-            ippiFilterSobelVertSecondGetBufferSize_8u16s_C1R(ippiSize(src.cols, src.rows), (IppiMaskSize)(ksize*10+ksize),&bufSize);
-            buffer.allocate(bufSize);
+            if ((dx == 0) && (dy == 2))
+            {
+                if (0 > ippiFilterSobelHorizSecondGetBufferSize_8u16s_C1R(ippiSize(src.cols, src.rows), (IppiMaskSize)(ksize*10+ksize),&bufSize))
+                    return false;
+                buffer.allocate(bufSize);
 
-            ippiFilterSobelVertSecondBorder_8u16s_C1R((const Ipp8u*)src.data, (int)src.step,
-               (Ipp16s*)dst.data, (int)dst.step, ippiSize(src.cols, src.rows), (IppiMaskSize)(ksize*10+ksize),
-                                      ippBorderRepl, 0, (Ipp8u*)(char*)buffer);
+                return (0 <= ippiFilterSobelHorizSecondBorder_8u16s_C1R((const Ipp8u*)src.data, (int)src.step,
+                                    (Ipp16s*)dst.data, (int)dst.step, ippiSize(src.cols, src.rows), (IppiMaskSize)(ksize*10+ksize),
+                                    ippBorderRepl, 0, (Ipp8u*)(char*)buffer));
+            }
+        }
 
-            return true;
-         }
+        if (src.type() == CV_32F && dst.type() == CV_32F)
+        {
+            if ((dx == 1) && (dy == 0))
+            {
+                if (0 > ippiFilterSobelNegVertGetBufferSize_32f_C1R(ippiSize(src.cols, src.rows), (IppiMaskSize)(ksize*10+ksize), &bufSize))
+                    return false;
+                buffer.allocate(bufSize);
 
-         if((dx == 0) && (dy == 2))
-         {
-            ippiFilterSobelHorizSecondGetBufferSize_8u16s_C1R(ippiSize(src.cols, src.rows), (IppiMaskSize)(ksize*10+ksize),&bufSize);
-            buffer.allocate(bufSize);
+                if (0 > ippiFilterSobelNegVertBorder_32f_C1R((const Ipp32f*)src.data, (int)src.step,
+                                (Ipp32f*)dst.data, (int)dst.step, ippiSize(src.cols, src.rows), (IppiMaskSize)(ksize*10+ksize),
+                                ippBorderRepl, 0, (Ipp8u*)(char*)buffer))
+                {
+                    return false;
+                }
+                if(scale != 1)
+                    ippiMulC_32f_C1R((Ipp32f *)dst.data, (int)dst.step, (Ipp32f)scale, (Ipp32f *)dst.data, (int)dst.step, ippiSize(dst.cols*dst.channels(), dst.rows));
+                return true;
+            }
 
-            ippiFilterSobelHorizSecondBorder_8u16s_C1R((const Ipp8u*)src.data, (int)src.step,
-               (Ipp16s*)dst.data, (int)dst.step, ippiSize(src.cols, src.rows), (IppiMaskSize)(ksize*10+ksize),
-                                      ippBorderRepl, 0, (Ipp8u*)(char*)buffer);
+            if ((dx == 0) && (dy == 1))
+            {
+                if (0 > ippiFilterSobelHorizGetBufferSize_32f_C1R(ippiSize(src.cols, src.rows), (IppiMaskSize)(ksize*10+ksize),&bufSize))
+                    return false;
+                buffer.allocate(bufSize);
 
-            return true;
-         }
-      }
+                if (0 > ippiFilterSobelHorizBorder_32f_C1R((const Ipp32f*)src.data, (int)src.step,
+                                (Ipp32f*)dst.data, (int)dst.step, ippiSize(src.cols, src.rows), (IppiMaskSize)(ksize*10+ksize),
+                                ippBorderRepl, 0, (Ipp8u*)(char*)buffer))
+                {
+                    return false;
+                }
+                if(scale != 1)
+                    ippiMulC_32f_C1R((Ipp32f *)dst.data, (int)dst.step, (Ipp32f)scale, (Ipp32f *)dst.data, (int)dst.step, ippiSize(dst.cols*dst.channels(), dst.rows));
+                return true;
+            }
 
-      if(src.type() == CV_32F && dst.type() == CV_32F)
-      {
-         if((dx == 1) && (dy == 0))
-         {
-            ippiFilterSobelNegVertGetBufferSize_32f_C1R(ippiSize(src.cols, src.rows), (IppiMaskSize)(ksize*10+ksize), &bufSize);
-            buffer.allocate(bufSize);
+            if((dx == 2) && (dy == 0))
+            {
+                if (0 > ippiFilterSobelVertSecondGetBufferSize_32f_C1R(ippiSize(src.cols, src.rows), (IppiMaskSize)(ksize*10+ksize),&bufSize))
+                    return false;
+                buffer.allocate(bufSize);
 
-            ippiFilterSobelNegVertBorder_32f_C1R((const Ipp32f*)src.data, (int)src.step,
-               (Ipp32f*)dst.data, (int)dst.step, ippiSize(src.cols, src.rows), (IppiMaskSize)(ksize*10+ksize),
-                                      ippBorderRepl, 0, (Ipp8u*)(char*)buffer);
-            if(scale != 1)
-               ippiMulC_32f_C1IR((Ipp32f)scale, (Ipp32f *)dst.data, (int)dst.step, ippiSize(dst.cols*dst.channels(), dst.rows));
+                if (0 > ippiFilterSobelVertSecondBorder_32f_C1R((const Ipp32f*)src.data, (int)src.step,
+                                (Ipp32f*)dst.data, (int)dst.step, ippiSize(src.cols, src.rows), (IppiMaskSize)(ksize*10+ksize),
+                                ippBorderRepl, 0, (Ipp8u*)(char*)buffer))
+                {
+                    return false;
+                }
+                if(scale != 1)
+                    ippiMulC_32f_C1R((Ipp32f *)dst.data, (int)dst.step, (Ipp32f)scale, (Ipp32f *)dst.data, (int)dst.step, ippiSize(dst.cols*dst.channels(), dst.rows));
+                return true;
+            }
 
-            return true;
-         }
+            if((dx == 0) && (dy == 2))
+            {
+                if (0 > ippiFilterSobelHorizSecondGetBufferSize_32f_C1R(ippiSize(src.cols, src.rows), (IppiMaskSize)(ksize*10+ksize),&bufSize))
+                    return false;
+                buffer.allocate(bufSize);
 
-         if((dx == 0) && (dy == 1))
-         {
-            ippiFilterSobelHorizGetBufferSize_32f_C1R(ippiSize(src.cols, src.rows), (IppiMaskSize)(ksize*10+ksize),&bufSize);
-            buffer.allocate(bufSize);
+                if (0 > ippiFilterSobelHorizSecondBorder_32f_C1R((const Ipp32f*)src.data, (int)src.step,
+                                (Ipp32f*)dst.data, (int)dst.step, ippiSize(src.cols, src.rows), (IppiMaskSize)(ksize*10+ksize),
+                                ippBorderRepl, 0, (Ipp8u*)(char*)buffer))
+                {
+                    return false;
+                }
 
-            ippiFilterSobelHorizBorder_32f_C1R((const Ipp32f*)src.data, (int)src.step,
-               (Ipp32f*)dst.data, (int)dst.step, ippiSize(src.cols, src.rows), (IppiMaskSize)(ksize*10+ksize),
-                                      ippBorderRepl, 0, (Ipp8u*)(char*)buffer);
-            if(scale != 1)
-               ippiMulC_32f_C1IR((Ipp32f)scale, (Ipp32f *)dst.data, (int)dst.step, ippiSize(dst.cols*dst.channels(), dst.rows));
+                if(scale != 1)
+                    ippiMulC_32f_C1R((Ipp32f *)dst.data, (int)dst.step, (Ipp32f)scale, (Ipp32f *)dst.data, (int)dst.step, ippiSize(dst.cols*dst.channels(), dst.rows));
+                return true;
+            }
+        }
+    }
 
-            return true;
-         }
-
-         if((dx == 2) && (dy == 0))
-         {
-            ippiFilterSobelVertSecondGetBufferSize_32f_C1R(ippiSize(src.cols, src.rows), (IppiMaskSize)(ksize*10+ksize),&bufSize);
-            buffer.allocate(bufSize);
-
-            ippiFilterSobelVertSecondBorder_32f_C1R((const Ipp32f*)src.data, (int)src.step,
-               (Ipp32f*)dst.data, (int)dst.step, ippiSize(src.cols, src.rows), (IppiMaskSize)(ksize*10+ksize),
-                                      ippBorderRepl, 0, (Ipp8u*)(char*)buffer);
-            if(scale != 1)
-               ippiMulC_32f_C1IR((Ipp32f)scale, (Ipp32f *)dst.data, (int)dst.step, ippiSize(dst.cols*dst.channels(), dst.rows));
-
-            return true;
-         }
-
-         if((dx == 0) && (dy == 2))
-         {
-            ippiFilterSobelHorizSecondGetBufferSize_32f_C1R(ippiSize(src.cols, src.rows), (IppiMaskSize)(ksize*10+ksize),&bufSize);
-            buffer.allocate(bufSize);
-
-            ippiFilterSobelHorizSecondBorder_32f_C1R((const Ipp32f*)src.data, (int)src.step,
-               (Ipp32f*)dst.data, (int)dst.step, ippiSize(src.cols, src.rows), (IppiMaskSize)(ksize*10+ksize),
-                                      ippBorderRepl, 0, (Ipp8u*)(char*)buffer);
-            if(scale != 1)
-               ippiMulC_32f_C1IR((Ipp32f)scale, (Ipp32f *)dst.data, (int)dst.step, ippiSize(dst.cols*dst.channels(), dst.rows));
-
-            return true;
-         }
-      }
-   }
-
-   if(ksize <= 0)
-      return IPPDerivScharr(src, dst, ddepth, dx, dy, scale);
-
-   return false;
+    if(ksize <= 0)
+        return IPPDerivScharr(src, dst, ddepth, dx, dy, scale);
+    return false;
 }
 
 }
@@ -436,7 +444,7 @@ void cv::Sobel( InputArray _src, OutputArray _dst, int ddepth, int dx, int dy,
     if(dx < 3 && dy < 3 && cn == 1 && borderType == BORDER_REPLICATE)
     {
         Mat src = _src.getMat(), dst = _dst.getMat();
-        if(IPPDeriv(src, dst, ddepth, dx, dy, ksize,scale))
+        if (IPPDeriv(src, dst, ddepth, dx, dy, ksize,scale))
             return;
     }
 #endif
diff --git a/modules/imgproc/src/filter.cpp b/modules/imgproc/src/filter.cpp
index 2bc6b8a70..d23f87ef4 100644
--- a/modules/imgproc/src/filter.cpp
+++ b/modules/imgproc/src/filter.cpp
@@ -1420,36 +1420,16 @@ struct RowVec_32f
 
     int operator()(const uchar* _src, uchar* _dst, int width, int cn) const
     {
+#ifdef USE_IPP_SEP_FILTERS
+        int ret = ippiOperator(_src, _dst, width, cn);
+        if (ret > 0)
+            return ret;
+#endif
         int _ksize = kernel.rows + kernel.cols - 1;
         const float* src0 = (const float*)_src;
         float* dst = (float*)_dst;
         const float* _kx = (const float*)kernel.data;
 
-#ifdef USE_IPP_SEP_FILTERS
-        IppiSize roisz = { width, 1 };
-        if( (cn == 1 || cn == 3) && width >= _ksize*8 )
-        {
-            if( bufsz < 0 )
-            {
-                if( (cn == 1 && ippiFilterRowBorderPipelineGetBufferSize_32f_C1R(roisz, _ksize, &bufsz) < 0) ||
-                    (cn == 3 && ippiFilterRowBorderPipelineGetBufferSize_32f_C3R(roisz, _ksize, &bufsz) < 0))
-                    return 0;
-            }
-            AutoBuffer<uchar> buf(bufsz + 64);
-            uchar* bufptr = alignPtr((uchar*)buf, 32);
-            int step = (int)(width*sizeof(dst[0])*cn);
-            float borderValue[] = {0.f, 0.f, 0.f};
-            // here is the trick. IPP needs border type and extrapolates the row. We did it already.
-            // So we pass anchor=0 and ignore the right tail of results since they are incorrect there.
-            if( (cn == 1 && ippiFilterRowBorderPipeline_32f_C1R(src0, step, &dst, roisz, _kx, _ksize, 0,
-                                                                ippBorderRepl, borderValue[0], bufptr) < 0) ||
-                (cn == 3 && ippiFilterRowBorderPipeline_32f_C3R(src0, step, &dst, roisz, _kx, _ksize, 0,
-                                                                ippBorderRepl, borderValue, bufptr) < 0))
-                return 0;
-            return width - _ksize + 1;
-        }
-#endif
-
         if( !haveSSE )
             return 0;
 
@@ -1479,7 +1459,38 @@ struct RowVec_32f
     Mat kernel;
     bool haveSSE;
 #ifdef USE_IPP_SEP_FILTERS
+private:
     mutable int bufsz;
+    int ippiOperator(const uchar* _src, uchar* _dst, int width, int cn) const
+    {
+        int _ksize = kernel.rows + kernel.cols - 1;
+        if ((1 != cn && 3 != cn) || width < _ksize*8)
+            return 0;
+
+        const float* src = (const float*)_src;
+        float* dst = (float*)_dst;
+        const float* _kx = (const float*)kernel.data;
+
+        IppiSize roisz = { width, 1 };
+        if( bufsz < 0 )
+        {
+            if( (cn == 1 && ippiFilterRowBorderPipelineGetBufferSize_32f_C1R(roisz, _ksize, &bufsz) < 0) ||
+                (cn == 3 && ippiFilterRowBorderPipelineGetBufferSize_32f_C3R(roisz, _ksize, &bufsz) < 0))
+                return 0;
+        }
+        AutoBuffer<uchar> buf(bufsz + 64);
+        uchar* bufptr = alignPtr((uchar*)buf, 32);
+        int step = (int)(width*sizeof(dst[0])*cn);
+        float borderValue[] = {0.f, 0.f, 0.f};
+        // here is the trick. IPP needs border type and extrapolates the row. We did it already.
+        // So we pass anchor=0 and ignore the right tail of results since they are incorrect there.
+        if( (cn == 1 && ippiFilterRowBorderPipeline_32f_C1R(src, step, &dst, roisz, _kx, _ksize, 0,
+                                                            ippBorderRepl, borderValue[0], bufptr) < 0) ||
+            (cn == 3 && ippiFilterRowBorderPipeline_32f_C3R(src, step, &dst, roisz, _kx, _ksize, 0,
+                                                            ippBorderRepl, borderValue, bufptr) < 0))
+            return 0;
+        return width - _ksize + 1;
+    }
 #endif
 };
 
diff --git a/modules/imgproc/src/imgwarp.cpp b/modules/imgproc/src/imgwarp.cpp
index 9d3eabaad..9e5540a7b 100644
--- a/modules/imgproc/src/imgwarp.cpp
+++ b/modules/imgproc/src/imgwarp.cpp
@@ -1971,7 +1971,7 @@ public:
         CHECK_IPP_STATUS(getBufferSizeFunc(pSpec, dstSize, cn, &bufsize));
         CHECK_IPP_STATUS(getSrcOffsetFunc(pSpec, dstOffset, &srcOffset));
 
-        Ipp8u* pSrc = (Ipp8u*)src.data + (int)src.step[0] * srcOffset.y + srcOffset.x * cn * itemSize;
+        const Ipp8u* pSrc = (const Ipp8u*)src.data + (int)src.step[0] * srcOffset.y + srcOffset.x * cn * itemSize;
         Ipp8u* pDst = (Ipp8u*)dst.data + (int)dst.step[0] * dstOffset.y + dstOffset.x * cn * itemSize;
 
         AutoBuffer<uchar> buf(bufsize + 64);
@@ -1980,7 +1980,6 @@ public:
         if( func( pSrc, (int)src.step[0], pDst, (int)dst.step[0], dstOffset, dstSize, ippBorderRepl, 0, pSpec, bufptr ) < 0 )
             *ok = false;
     }
-
 private:
     const Mat & src;
     Mat & dst;
@@ -4025,25 +4024,25 @@ public:
           *ok = true;
       }
 
-      virtual void operator() (const Range& range) const
-      {
-          IppiSize srcsize = { src.cols, src.rows };
-          IppiRect srcroi = { 0, 0, src.cols, src.rows };
-          IppiRect dstroi = { 0, range.start, dst.cols, range.end - range.start };
-          int cnn = src.channels();
-          if( borderType == BORDER_CONSTANT )
-          {
-              IppiSize setSize = { dst.cols, range.end - range.start };
-              void *dataPointer = dst.data + dst.step[0] * range.start;
-              if( !IPPSet( borderValue, dataPointer, (int)dst.step[0], setSize, cnn, src.depth() ) )
-              {
-                  *ok = false;
-                  return;
-              }
-          }
-          if( func( src.data, srcsize, (int)src.step[0], srcroi, dst.data, (int)dst.step[0], dstroi, coeffs, mode ) < 0) ////Aug 2013: problem in IPP 7.1, 8.0 : sometimes function return ippStsCoeffErr
-              *ok = false;
-      }
+    virtual void operator() (const Range& range) const
+    {
+        IppiSize srcsize = { src.cols, src.rows };
+        IppiRect srcroi = { 0, 0, src.cols, src.rows };
+        IppiRect dstroi = { 0, range.start, dst.cols, range.end - range.start };
+        int cnn = src.channels();
+        if( borderType == BORDER_CONSTANT )
+        {
+            IppiSize setSize = { dst.cols, range.end - range.start };
+            void *dataPointer = dst.data + dst.step[0] * range.start;
+            if( !IPPSet( borderValue, dataPointer, (int)dst.step[0], setSize, cnn, src.depth() ) )
+            {
+                *ok = false;
+                return;
+            }
+        }
+        if( func( src.data, srcsize, (int)src.step[0], srcroi, dst.data, (int)dst.step[0], dstroi, coeffs, mode ) < 0) ////Aug 2013: problem in IPP 7.1, 8.0 : sometimes function return ippStsCoeffErr
+            *ok = false;
+    }
 private:
     Mat &src;
     Mat &dst;
@@ -4368,26 +4367,26 @@ public:
           *ok = true;
       }
 
-      virtual void operator() (const Range& range) const
-      {
-          IppiSize srcsize = {src.cols, src.rows};
-          IppiRect srcroi = {0, 0, src.cols, src.rows};
-          IppiRect dstroi = {0, range.start, dst.cols, range.end - range.start};
-          int cnn = src.channels();
+    virtual void operator() (const Range& range) const
+    {
+        IppiSize srcsize = {src.cols, src.rows};
+        IppiRect srcroi = {0, 0, src.cols, src.rows};
+        IppiRect dstroi = {0, range.start, dst.cols, range.end - range.start};
+        int cnn = src.channels();
 
-          if( borderType == BORDER_CONSTANT )
-          {
-              IppiSize setSize = {dst.cols, range.end - range.start};
-              void *dataPointer = dst.data + dst.step[0] * range.start;
-              if( !IPPSet( borderValue, dataPointer, (int)dst.step[0], setSize, cnn, src.depth() ) )
-              {
-                  *ok = false;
-                  return;
-              }
-          }
-          if( func(src.data, srcsize, (int)src.step[0], srcroi, dst.data, (int)dst.step[0], dstroi, coeffs, mode) < 0)
-              *ok = false;
-      }
+        if( borderType == BORDER_CONSTANT )
+        {
+            IppiSize setSize = {dst.cols, range.end - range.start};
+            void *dataPointer = dst.data + dst.step[0] * range.start;
+            if( !IPPSet( borderValue, dataPointer, (int)dst.step[0], setSize, cnn, src.depth() ) )
+            {
+                *ok = false;
+                return;
+            }
+        }
+        if( func(src.data, srcsize, (int)src.step[0], srcroi, dst.data, (int)dst.step[0], dstroi, coeffs, mode) < 0)
+            *ok = false;
+    }
 private:
     Mat &src;
     Mat &dst;
diff --git a/modules/imgproc/src/morph.cpp b/modules/imgproc/src/morph.cpp
index 1dd0a252e..b3433549d 100644
--- a/modules/imgproc/src/morph.cpp
+++ b/modules/imgproc/src/morph.cpp
@@ -1136,80 +1136,128 @@ private:
     Scalar borderValue;
 };
 
-#if defined (HAVE_IPP) && (IPP_VERSION_MAJOR >= 7)
+#if defined (HAVE_IPP) && (IPP_VERSION_MAJOR >= 8) && (IPP_VERSION_MINOR >= 1)
 static bool IPPMorphReplicate(int op, const Mat &src, Mat &dst, const Mat &kernel,
                               const Size& ksize, const Point &anchor, bool rectKernel)
 {
     int type = src.type();
     const Mat* _src = &src;
     Mat temp;
-    if( src.data == dst.data )
+    if (src.data == dst.data)
     {
         src.copyTo(temp);
         _src = &temp;
     }
-    //DEPRECATED. Allocates and initializes morphology state structure for erosion or dilation operation.
-    typedef IppStatus (CV_STDCALL* ippiMorphologyInitAllocFunc)(int, const void*, IppiSize, IppiPoint, IppiMorphState **);
-    typedef IppStatus (CV_STDCALL* ippiMorphologyBorderReplicateFunc)(const void*, int, void *, int,
-                                                                      IppiSize, IppiBorderType, IppiMorphState *);
-    typedef IppStatus (CV_STDCALL* ippiFilterMinMaxGetBufferSizeFunc)(int, IppiSize, int*);
-    typedef IppStatus (CV_STDCALL* ippiFilterMinMaxBorderReplicateFunc)(const void*, int, void*, int,
-                                                                        IppiSize, IppiSize, IppiPoint, void*);
-
-    ippiMorphologyInitAllocFunc initAllocFunc = 0;
-    ippiMorphologyBorderReplicateFunc morphFunc = 0;
-    ippiFilterMinMaxGetBufferSizeFunc getBufSizeFunc = 0;
-    ippiFilterMinMaxBorderReplicateFunc morphRectFunc = 0;
-
-    #define IPP_MORPH_CASE(type, flavor) \
-    case type: \
-        initAllocFunc = (ippiMorphologyInitAllocFunc)ippiMorphologyInitAlloc_##flavor; \
-        morphFunc = op == MORPH_ERODE ? (ippiMorphologyBorderReplicateFunc)ippiErodeBorderReplicate_##flavor : \
-                                        (ippiMorphologyBorderReplicateFunc)ippiDilateBorderReplicate_##flavor; \
-        getBufSizeFunc = (ippiFilterMinMaxGetBufferSizeFunc)ippiFilterMinGetBufferSize_##flavor; \
-        morphRectFunc = op == MORPH_ERODE ? (ippiFilterMinMaxBorderReplicateFunc)ippiFilterMinBorderReplicate_##flavor : \
-                                            (ippiFilterMinMaxBorderReplicateFunc)ippiFilterMaxBorderReplicate_##flavor; \
-        break
-
-    switch( type )
-    {
-    IPP_MORPH_CASE(CV_8UC1, 8u_C1R);
-    IPP_MORPH_CASE(CV_8UC3, 8u_C3R);
-    IPP_MORPH_CASE(CV_8UC4, 8u_C4R);
-    IPP_MORPH_CASE(CV_32FC1, 32f_C1R);
-    IPP_MORPH_CASE(CV_32FC3, 32f_C3R);
-    IPP_MORPH_CASE(CV_32FC4, 32f_C4R);
-    default:
-        return false;
-    }
-    #undef IPP_MORPH_CASE
 
     IppiSize roiSize = {src.cols, src.rows};
     IppiSize kernelSize = {ksize.width, ksize.height};
-    IppiPoint point = {anchor.x, anchor.y};
 
-    if( !rectKernel && morphFunc && initAllocFunc )
+    if (!rectKernel)
     {
-        IppiMorphState* pState;
-        if( initAllocFunc( roiSize.width, kernel.data, kernelSize, point, &pState ) < 0 )
+#if 1
+        if (((kernel.cols - 1) / 2 != anchor.x) || ((kernel.rows - 1) / 2 != anchor.y))
             return false;
-        bool is_ok = morphFunc( _src->data, (int)_src->step[0],
-                               dst.data, (int)dst.step[0],
-                               roiSize, ippBorderRepl, pState ) >= 0;
-        ippiMorphologyFree(pState);
-        return is_ok;
+        #define IPP_MORPH_CASE(cvtype, flavor, data_type) \
+        case cvtype: \
+            {\
+                int specSize = 0, bufferSize = 0;\
+                if (0 > ippiMorphologyBorderGetSize_##flavor(roiSize.width, kernelSize, &specSize, &bufferSize))\
+                    return false;\
+                IppiMorphState *pSpec = (IppiMorphState*)ippMalloc(specSize);\
+                Ipp8u *pBuffer = (Ipp8u*)ippMalloc(bufferSize);\
+                if (0 > ippiMorphologyBorderInit_##flavor(roiSize.width, kernel.data, kernelSize, pSpec, pBuffer))\
+                {\
+                    ippFree(pBuffer);\
+                    ippFree(pSpec);\
+                    return false;\
+                }\
+                bool ok = false;\
+                if (op == MORPH_ERODE)\
+                    ok = (0 <= ippiErodeBorder_##flavor((Ipp##data_type *)_src->data, (int)_src->step[0], (Ipp##data_type *)dst.data, (int)dst.step[0],\
+                                            roiSize, ippBorderRepl, 0, pSpec, pBuffer));\
+                else\
+                    ok = (0 <= ippiDilateBorder_##flavor((Ipp##data_type *)_src->data, (int)_src->step[0], (Ipp##data_type *)dst.data, (int)dst.step[0],\
+                                            roiSize, ippBorderRepl, 0, pSpec, pBuffer));\
+                ippFree(pBuffer);\
+                ippFree(pSpec);\
+                return ok;\
+            }\
+            break;
+#else
+        IppiPoint point = {anchor.x, anchor.y};
+        // this is case, which can be used with the anchor not in center of the kernel, but
+        // ippiMorphologyBorderGetSize_, ippiErodeBorderReplicate_ and ippiDilateBorderReplicate_ are deprecated.
+        #define IPP_MORPH_CASE(cvtype, flavor, data_type) \
+        case cvtype: \
+            {\
+                int specSize = 0;\
+                int bufferSize = 0;\
+                if (0 > ippiMorphologyGetSize_##flavor( roiSize.width, kernel.data kernelSize, &specSize))\
+                    return false;\
+                bool ok = false;\
+                IppiMorphState* pState = (IppiMorphState*)ippMalloc(specSize);\
+                if (ippiMorphologyInit_##flavor(roiSize.width, kernel.data, kernelSize, point, pState) >= 0)\
+                {\
+                    if (op == MORPH_ERODE)\
+                        ok = ippiErodeBorderReplicate_##flavor((Ipp##data_type *)_src->data, (int)_src->step[0],\
+                            (Ipp##data_type *)dst.data, (int)dst.step[0],\
+                            roiSize, ippBorderRepl, pState ) >= 0;\
+                    else\
+                        ok = ippiDilateBorderReplicate_##flavor((Ipp##data_type *)_src->data, (int)_src->step[0],\
+                            (Ipp##data_type *)dst.data, (int)dst.step[0],\
+                            roiSize, ippBorderRepl, pState ) >= 0;\
+                }\
+                ippFree(pState);\
+                return ok;\
+            }\
+            break;
+#endif
+        switch (type)
+        {
+        IPP_MORPH_CASE(CV_8UC1, 8u_C1R, 8u);
+        IPP_MORPH_CASE(CV_8UC3, 8u_C3R, 8u);
+        IPP_MORPH_CASE(CV_8UC4, 8u_C4R, 8u);
+        IPP_MORPH_CASE(CV_32FC1, 32f_C1R, 32f);
+        IPP_MORPH_CASE(CV_32FC3, 32f_C3R, 32f);
+        IPP_MORPH_CASE(CV_32FC4, 32f_C4R, 32f);
+        default:
+            return false;
+        }
+
+        #undef IPP_MORPH_CASE
     }
-    else if( rectKernel && morphRectFunc && getBufSizeFunc )
+    else
     {
-        int bufSize = 0;
-        if( getBufSizeFunc( src.cols, kernelSize, &bufSize) < 0 )
+        IppiPoint point = {anchor.x, anchor.y};
+
+        #define IPP_MORPH_CASE(cvtype, flavor, data_type) \
+        case cvtype: \
+            {\
+                int bufSize = 0;\
+                if (0 > ippiFilterMinGetBufferSize_##flavor(src.cols, kernelSize, &bufSize))\
+                    return false;\
+                AutoBuffer<uchar> buf(bufSize + 64);\
+                uchar* buffer = alignPtr((uchar*)buf, 32);\
+                if (op == MORPH_ERODE)\
+                    return (0 <= ippiFilterMinBorderReplicate_##flavor((Ipp##data_type *)_src->data, (int)_src->step[0], (Ipp##data_type *)dst.data, (int)dst.step[0], roiSize, kernelSize, point, buffer));\
+                return (0 <= ippiFilterMaxBorderReplicate_##flavor((Ipp##data_type *)_src->data, (int)_src->step[0], (Ipp##data_type *)dst.data, (int)dst.step[0], roiSize, kernelSize, point, buffer));\
+            }\
+            break;
+
+        switch (type)
+        {
+        IPP_MORPH_CASE(CV_8UC1, 8u_C1R, 8u);
+        IPP_MORPH_CASE(CV_8UC3, 8u_C3R, 8u);
+        IPP_MORPH_CASE(CV_8UC4, 8u_C4R, 8u);
+        IPP_MORPH_CASE(CV_32FC1, 32f_C1R, 32f);
+        IPP_MORPH_CASE(CV_32FC3, 32f_C3R, 32f);
+        IPP_MORPH_CASE(CV_32FC4, 32f_C4R, 32f);
+        default:
             return false;
-        AutoBuffer<uchar> buf(bufSize + 64);
-        uchar* buffer = alignPtr((uchar*)buf, 32);
-        return morphRectFunc(_src->data, (int)_src->step[0], dst.data, (int)dst.step[0],
-                             roiSize, kernelSize, point, buffer) >= 0;
+        }
+
+        #undef IPP_MORPH_CASE
     }
-    return false;
 }
 
 static bool IPPMorphOp(int op, InputArray _src, OutputArray _dst,
@@ -1411,7 +1459,7 @@ static void morphOp( int op, InputArray _src, OutputArray _dst,
     Size ksize = kernel.data ? kernel.size() : Size(3,3);
     anchor = normalizeAnchor(anchor, ksize);
 
-#if defined (HAVE_IPP) && (IPP_VERSION_MAJOR >= 7)
+#if defined (HAVE_IPP) && (IPP_VERSION_MAJOR >= 8) && (IPP_VERSION_MINOR >= 1)
     if( IPPMorphOp(op, _src, _dst, kernel, anchor, iterations, borderType, borderValue) )
         return;
 #endif
diff --git a/modules/imgproc/src/smooth.cpp b/modules/imgproc/src/smooth.cpp
index 6a18af5c2..7499e2468 100644
--- a/modules/imgproc/src/smooth.cpp
+++ b/modules/imgproc/src/smooth.cpp
@@ -1109,20 +1109,27 @@ void cv::GaussianBlur( InputArray _src, OutputArray _dst, Size ksize,
         return;
 #endif
 
-#if defined HAVE_IPP && (IPP_VERSION_MAJOR >= 7)
+#if defined (HAVE_IPP) && (IPP_VERSION_MAJOR >= 8) && (IPP_VERSION_MINOR >= 1)
     if( type == CV_32FC1 && sigma1 == sigma2 && ksize.width == ksize.height && sigma1 != 0.0 )
     {
         Mat src = _src.getMat(), dst = _dst.getMat();
         IppiSize roi = { src.cols, src.rows };
-        int bufSize = 0;
-        ippiFilterGaussGetBufferSize_32f_C1R(roi, ksize.width, &bufSize);
-        AutoBuffer<uchar> buf(bufSize+128);
-        if( ippiFilterGaussBorder_32f_C1R((const Ipp32f *)src.data, (int)src.step,
-                                          (Ipp32f *)dst.data, (int)dst.step,
-                                          roi, ksize.width, (Ipp32f)sigma1,
-                                          (IppiBorderType)borderType, 0.0,
-                                          alignPtr(&buf[0],32)) >= 0 )
-            return;
+        int specSize = 0, bufferSize = 0;
+        if (0 <=  ippiFilterGaussianGetBufferSize(roi, (Ipp32u)ksize.width, ipp32f, 1, &specSize, &bufferSize))
+        {
+            IppFilterGaussianSpec *pSpec = (IppFilterGaussianSpec*)ippMalloc(specSize);
+            Ipp8u *pBuffer = (Ipp8u*)ippMalloc(bufferSize);
+            if (0 <= ippiFilterGaussianInit(roi, (Ipp32u)ksize.width, (Ipp32f)sigma1, (IppiBorderType)borderType, ipp32f, 1, pSpec, pBuffer))
+            {
+                IppStatus sts = ippiFilterGaussianBorder_32f_C1R( (const Ipp32f *)src.data, (int)src.step,
+                                                                     (Ipp32f *)dst.data, (int)dst.step,
+                                                                     roi,  0.0, pSpec, pBuffer);
+                ippFree(pBuffer);
+                ippFree(pSpec);
+                if (0 <= sts)
+                    return;
+            }
+        }
     }
 #endif
 
@@ -2180,11 +2187,19 @@ public:
           IppiSize kernel = {d, d};
           IppiSize roi={dst.cols, range.end - range.start};
           int bufsize=0;
-          ippiFilterBilateralGetBufSize_8u_C1R( ippiFilterBilateralGauss, roi, kernel, &bufsize);
+          if (0 > ippiFilterBilateralGetBufSize_8u_C1R( ippiFilterBilateralGauss, roi, kernel, &bufsize))
+          {
+              *ok = false;
+              return;
+          }
           AutoBuffer<uchar> buf(bufsize);
           IppiFilterBilateralSpec *pSpec = (IppiFilterBilateralSpec *)alignPtr(&buf[0], 32);
-          ippiFilterBilateralInit_8u_C1R( ippiFilterBilateralGauss, kernel, (Ipp32f)sigma_color, (Ipp32f)sigma_space, 1, pSpec );
-          if( ippiFilterBilateral_8u_C1R( src.ptr<uchar>(range.start) + radius * ((int)src.step[0] + 1), (int)src.step[0], dst.ptr<uchar>(range.start), (int)dst.step[0], roi, kernel, pSpec ) < 0)
+          if (0 > ippiFilterBilateralInit_8u_C1R( ippiFilterBilateralGauss, kernel, (Ipp32f)sigma_color, (Ipp32f)sigma_space, 1, pSpec ))
+          {
+              *ok = false;
+              return;
+          }
+          if (0 > ippiFilterBilateral_8u_C1R( src.ptr<uchar>(range.start) + radius * ((int)src.step[0] + 1), (int)src.step[0], dst.ptr<uchar>(range.start), (int)dst.step[0], roi, kernel, pSpec ))
               *ok = false;
       }
 private:
diff --git a/modules/imgproc/src/sumpixels.cpp b/modules/imgproc/src/sumpixels.cpp
index 4e18f119f..c32813fab 100644
--- a/modules/imgproc/src/sumpixels.cpp
+++ b/modules/imgproc/src/sumpixels.cpp
@@ -365,30 +365,32 @@ void cv::integral( InputArray _src, OutputArray _sum, OutputArray _sqsum, Output
 #if defined (HAVE_IPP) && (IPP_VERSION_MAJOR >= 7)
     if( ( depth == CV_8U ) && ( sdepth == CV_32F || sdepth == CV_32S ) && ( !_tilted.needed() ) && ( !_sqsum.needed() || sqdepth == CV_64F ) && ( cn == 1 ) )
     {
+        IppStatus status = ippStsErr;
         IppiSize srcRoiSize = ippiSize( src.cols, src.rows );
         if( sdepth == CV_32F )
         {
             if( _sqsum.needed() )
             {
-                ippiSqrIntegral_8u32f64f_C1R( (const Ipp8u*)src.data, (int)src.step, (Ipp32f*)sum.data, (int)sum.step, (Ipp64f*)sqsum.data, (int)sqsum.step, srcRoiSize, 0, 0 );
+                status = ippiSqrIntegral_8u32f64f_C1R( (const Ipp8u*)src.data, (int)src.step, (Ipp32f*)sum.data, (int)sum.step, (Ipp64f*)sqsum.data, (int)sqsum.step, srcRoiSize, 0, 0 );
             }
             else
             {
-                ippiIntegral_8u32f_C1R( (const Ipp8u*)src.data, (int)src.step, (Ipp32f*)sum.data, (int)sum.step, srcRoiSize, 0 );
+                status = ippiIntegral_8u32f_C1R( (const Ipp8u*)src.data, (int)src.step, (Ipp32f*)sum.data, (int)sum.step, srcRoiSize, 0 );
             }
         }
         else if( sdepth == CV_32S )
         {
             if( _sqsum.needed() )
             {
-                ippiSqrIntegral_8u32s64f_C1R( (const Ipp8u*)src.data, (int)src.step, (Ipp32s*)sum.data, (int)sum.step, (Ipp64f*)sqsum.data, (int)sqsum.step, srcRoiSize, 0, 0 );
+                status = ippiSqrIntegral_8u32s64f_C1R( (const Ipp8u*)src.data, (int)src.step, (Ipp32s*)sum.data, (int)sum.step, (Ipp64f*)sqsum.data, (int)sqsum.step, srcRoiSize, 0, 0 );
             }
             else
             {
-                ippiIntegral_8u32s_C1R( (const Ipp8u*)src.data, (int)src.step, (Ipp32s*)sum.data, (int)sum.step, srcRoiSize, 0 );
+                status = ippiIntegral_8u32s_C1R( (const Ipp8u*)src.data, (int)src.step, (Ipp32s*)sum.data, (int)sum.step, srcRoiSize, 0 );
             }
         }
-        return;
+        if (0 <= status)
+            return;
     }
 #endif
 
diff --git a/modules/objdetect/src/haar.cpp b/modules/objdetect/src/haar.cpp
index 2f864797f..4b2d2b0d5 100644
--- a/modules/objdetect/src/haar.cpp
+++ b/modules/objdetect/src/haar.cpp
@@ -1318,9 +1318,9 @@ public:
         if( cascade->hid_cascade->ipp_stages )
         {
             IppiRect iequRect = {equRect.x, equRect.y, equRect.width, equRect.height};
-            ippiRectStdDev_32f_C1R(sum1.ptr<float>(y1), sum1.step,
-                                   sqsum1.ptr<double>(y1), sqsum1.step,
-                                   norm1->ptr<float>(y1), norm1->step,
+            ippiRectStdDev_32f_C1R(sum1.ptr<float>(y1), (int)sum1.step,
+                                   sqsum1.ptr<double>(y1), (int)sqsum1.step,
+                                   norm1->ptr<float>(y1), (int)norm1->step,
                                    ippiSize(ssz.width, ssz.height), iequRect );
 
             int positive = (ssz.width/ystep)*((ssz.height + ystep-1)/ystep);
@@ -1341,9 +1341,9 @@ public:
             for( int j = 0; j < cascade->count; j++ )
             {
                 if( ippiApplyHaarClassifier_32f_C1R(
-                            sum1.ptr<float>(y1), sum1.step,
-                            norm1->ptr<float>(y1), norm1->step,
-                            mask1->ptr<uchar>(y1), mask1->step,
+                            sum1.ptr<float>(y1), (int)sum1.step,
+                            norm1->ptr<float>(y1), (int)norm1->step,
+                            mask1->ptr<uchar>(y1), (int)mask1->step,
                             ippiSize(ssz.width, ssz.height), &positive,
                             cascade->hid_cascade->stage_classifier[j].threshold,
                             (IppiHaarClassifier_32f*)cascade->hid_cascade->ipp_stages[j]) < 0 )