generic-library/opencv
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Merge cuda-geek/soft-cascade-gpu into cuda-dev

Browse Source
This commit is contained in:
marina.kolpakova
2012-11-26 18:57:47 +04:00
parent 1712d0930c 05cd88ae42
commit 209f16455d
16 changed files with 2199 additions and 17 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
modules/gpu/src/cuda/bf_knnmatch.cu
View File

@@ -1034,7 +1034,7 @@ namespace cv { namespace gpu { namespace device
cudaSafeCall( cudaDeviceSynchronize() );
}
void findKnnMatchDispatcher(int k, const PtrStepSzb& trainIdx, const PtrStepSzb& distance, const PtrStepSzf& allDist, int cc, cudaStream_t stream)
void findKnnMatchDispatcher(int k, const PtrStepSzb& trainIdx, const PtrStepSzb& distance, const PtrStepSzf& allDist, int /*cc*/, cudaStream_t stream)
{
findKnnMatch<256>(k, static_cast<PtrStepSzi>(trainIdx), static_cast<PtrStepSzf>(distance), allDist, stream);
}

370
modules/gpu/src/cuda/icf-sc.cu Normal file
View File

@@ -0,0 +1,370 @@
/*M///////////////////////////////////////////////////////////////////////////////////////
//
// IMPORTANT: READ BEFORE DOWNLOADING, COPYING, INSTALLING OR USING.
//
// By downloading, copying, installing or using the software you agree to this license.
// If you do not agree to this license, do not download, install,
// copy or use the software.
//
//
// License Agreement
// For Open Source Computer Vision Library
//
// Copyright (C) 2000-2008, Intel Corporation, all rights reserved.
// Copyright (C) 2008-2012, Willow Garage Inc., all rights reserved.
// Third party copyrights are property of their respective owners.
//
// Redistribution and use in source and binary forms, with or without modification,
// are permitted provided that the following conditions are met:
//
// * Redistribution's of source code must retain the above copyright notice,
// this list of conditions and the following disclaimer.
//
// * Redistribution's in binary form must reproduce the above copyright notice,
// this list of conditions and the following disclaimer in the documentation
// and/or other materials provided with the distribution.
//
// * The name of the copyright holders may not be used to endorse or promote products
// derived from this software without specific prior written permission.
//
// This software is provided by the copyright holders and contributors "as is" and
// any express or implied warranties, including, but not limited to, the implied
// warranties of merchantability and fitness for a particular purpose are disclaimed.
// In no event shall the Intel Corporation or contributors be liable for any direct,
// indirect, incidental, special, exemplary, or consequential damages
// (including, but not limited to, procurement of substitute goods or services;
// loss of use, data, or profits; or business interruption) however caused
// and on any theory of liability, whether in contract, strict liability,
// or tort (including negligence or otherwise) arising in any way out of
// the use of this software, even if advised of the possibility of such damage.
//
//M*/
#include <opencv2/gpu/device/common.hpp>
#include <icf.hpp>
#include <float.h>
#include <stdio.h>
namespace cv { namespace gpu { namespace device {
namespace icf {
// ToDo: use textures or uncached load instruction.
__global__ void magToHist(const uchar* __restrict__ mag,
const float* __restrict__ angle, const int angPitch,
uchar* __restrict__ hog, const int hogPitch, const int fh)
{
const int y = blockIdx.y * blockDim.y + threadIdx.y;
const int x = blockIdx.x * blockDim.x + threadIdx.x;
const int bin = (int)(angle[y * angPitch + x]);
const uchar val = mag[y * hogPitch + x];
hog[((fh * bin) + y) * hogPitch + x] = val;
}
void fillBins(cv::gpu::PtrStepSzb hogluv, const cv::gpu::PtrStepSzf& nangle,
const int fw, const int fh, const int bins, cudaStream_t stream )
{
const uchar* mag = (const uchar*)hogluv.ptr(fh * bins);
uchar* hog = (uchar*)hogluv.ptr();
const float* angle = (const float*)nangle.ptr();
dim3 block(32, 8);
dim3 grid(fw / 32, fh / 8);
magToHist<<<grid, block, 0, stream>>>(mag, angle, nangle.step / sizeof(float), hog, hogluv.step, fh);
if (!stream)
{
cudaSafeCall( cudaGetLastError() );
cudaSafeCall( cudaDeviceSynchronize() );
}
}
__device__ __forceinline__ float overlapArea(const Detection &a, const Detection &b)
{
int w = ::min(a.x + a.w, b.x + b.w) - ::max(a.x, b.x);
int h = ::min(a.y + a.h, b.y + b.h) - ::max(a.y, b.y);
return (w < 0 || h < 0)? 0.f : (float)(w * h);
}
texture<uint4, cudaTextureType2D, cudaReadModeElementType> tdetections;
__global__ void overlap(const uint* n, uchar* overlaps)
{
const int idx = threadIdx.x;
const int total = *n;
for (int i = idx + 1; i < total; i += 192)
{
const uint4 _a = tex2D(tdetections, i, 0);
const Detection& a = *((Detection*)(&_a));
bool excluded = false;
for (int j = i + 1; j < total; ++j)
{
const uint4 _b = tex2D(tdetections, j, 0);
const Detection& b = *((Detection*)(&_b));
float ovl = overlapArea(a, b) / ::min(a.w * a.h, b.w * b.h);
if (ovl > 0.65f)
{
int suppessed = (a.confidence > b.confidence)? j : i;
overlaps[suppessed] = 1;
excluded = excluded || (suppessed == i);
}
if (__all(excluded)) break;
}
}
}
__global__ void collect(const uint* n, uchar* overlaps, uint* ctr, uint4* suppressed)
{
const int idx = threadIdx.x;
const int total = *n;
for (int i = idx; i < total; i += 192)
{
if (!overlaps[i])
{
int oidx = atomicInc(ctr, 50);
suppressed[oidx] = tex2D(tdetections, i + 1, 0);
}
}
}
void suppress(const PtrStepSzb& objects, PtrStepSzb overlaps, PtrStepSzi ndetections,
PtrStepSzb suppressed, cudaStream_t stream)
{
int block = 192;
int grid = 1;
cudaChannelFormatDesc desc = cudaCreateChannelDesc<uint4>();
size_t offset;
cudaSafeCall( cudaBindTexture2D(&offset, tdetections, objects.data, desc, objects.cols / sizeof(uint4), objects.rows, objects.step));
overlap<<<grid, block>>>((uint*)ndetections.ptr(0), (uchar*)overlaps.ptr(0));
collect<<<grid, block>>>((uint*)ndetections.ptr(0), (uchar*)overlaps.ptr(0), (uint*)suppressed.ptr(0), ((uint4*)suppressed.ptr(0)) + 1);
if (!stream)
{
cudaSafeCall( cudaGetLastError());
cudaSafeCall( cudaDeviceSynchronize());
}
}
template<typename Policy>
struct PrefixSum
{
__device static void apply(float& impact)
{
#if defined __CUDA_ARCH__ && __CUDA_ARCH__ >= 300
#pragma unroll
// scan on shuffl functions
for (int i = 1; i < Policy::WARP; i *= 2)
{
const float n = __shfl_up(impact, i, Policy::WARP);
if (threadIdx.x >= i)
impact += n;
}
#else
__shared__ volatile float ptr[Policy::STA_X * Policy::STA_Y];
const int idx = threadIdx.y * Policy::STA_X + threadIdx.x;
ptr[idx] = impact;
if ( threadIdx.x >= 1) ptr [idx ] = (ptr [idx - 1] + ptr [idx]);
if ( threadIdx.x >= 2) ptr [idx ] = (ptr [idx - 2] + ptr [idx]);
if ( threadIdx.x >= 4) ptr [idx ] = (ptr [idx - 4] + ptr [idx]);
if ( threadIdx.x >= 8) ptr [idx ] = (ptr [idx - 8] + ptr [idx]);
if ( threadIdx.x >= 16) ptr [idx ] = (ptr [idx - 16] + ptr [idx]);
impact = ptr[idx];
#endif
}
};
texture<int, cudaTextureType2D, cudaReadModeElementType> thogluv;
template<bool isUp>
__device__ __forceinline__ float rescale(const Level& level, Node& node)
{
uchar4& scaledRect = node.rect;
float relScale = level.relScale;
float farea = (scaledRect.z - scaledRect.x) * (scaledRect.w - scaledRect.y);
// rescale
scaledRect.x = __float2int_rn(relScale * scaledRect.x);
scaledRect.y = __float2int_rn(relScale * scaledRect.y);
scaledRect.z = __float2int_rn(relScale * scaledRect.z);
scaledRect.w = __float2int_rn(relScale * scaledRect.w);
float sarea = (scaledRect.z - scaledRect.x) * (scaledRect.w - scaledRect.y);
const float expected_new_area = farea * relScale * relScale;
float approx = (sarea == 0)? 1: __fdividef(sarea, expected_new_area);
float rootThreshold = (node.threshold & 0x0FFFFFFFU) * approx * level.scaling[(node.threshold >> 28) > 6];
return rootThreshold;
}
template<>
__device__ __forceinline__ float rescale<true>(const Level& level, Node& node)
{
uchar4& scaledRect = node.rect;
float relScale = level.relScale;
float farea = scaledRect.z * scaledRect.w;
// rescale
scaledRect.x = __float2int_rn(relScale * scaledRect.x);
scaledRect.y = __float2int_rn(relScale * scaledRect.y);
scaledRect.z = __float2int_rn(relScale * scaledRect.z);
scaledRect.w = __float2int_rn(relScale * scaledRect.w);
float sarea = scaledRect.z * scaledRect.w;
const float expected_new_area = farea * relScale * relScale;
float approx = __fdividef(sarea, expected_new_area);
float rootThreshold = (node.threshold & 0x0FFFFFFFU) * approx * level.scaling[(node.threshold >> 28) > 6];
return rootThreshold;
}
template<bool isUp>
__device__ __forceinline__ int get(int x, int y, uchar4 area)
{
int a = tex2D(thogluv, x + area.x, y + area.y);
int b = tex2D(thogluv, x + area.z, y + area.y);
int c = tex2D(thogluv, x + area.z, y + area.w);
int d = tex2D(thogluv, x + area.x, y + area.w);
return (a - b + c - d);
}
template<>
__device__ __forceinline__ int get<true>(int x, int y, uchar4 area)
{
x += area.x;
y += area.y;
int a = tex2D(thogluv, x, y);
int b = tex2D(thogluv, x + area.z, y);
int c = tex2D(thogluv, x + area.z, y + area.w);
int d = tex2D(thogluv, x, y + area.w);
return (a - b + c - d);
}
texture<float2, cudaTextureType2D, cudaReadModeElementType> troi;
template<typename Policy>
template<bool isUp>
__device void CascadeInvoker<Policy>::detect(Detection* objects, const uint ndetections, uint* ctr, const int downscales) const
{
const int y = blockIdx.y * blockDim.y + threadIdx.y;
const int x = blockIdx.x;
// load Lavel
__shared__ Level level;
// check POI
__shared__ volatile char roiCache[Policy::STA_Y];
if (!threadIdx.y && !threadIdx.x)
((float2*)roiCache)[threadIdx.x] = tex2D(troi, blockIdx.y, x);
__syncthreads();
if (!roiCache[threadIdx.y]) return;
if (!threadIdx.x)
level = levels[downscales + blockIdx.z];
if(x >= level.workRect.x || y >= level.workRect.y) return;
int st = level.octave * level.step;
const int stEnd = st + level.step;
const int hogluvStep = gridDim.y * Policy::STA_Y;
float confidence = 0.f;
for(; st < stEnd; st += Policy::WARP)
{
const int nId = (st + threadIdx.x) * 3;
Node node = nodes[nId];
float threshold = rescale<isUp>(level, node);
int sum = get<isUp>(x, y + (node.threshold >> 28) * hogluvStep, node.rect);
int next = 1 + (int)(sum >= threshold);
node = nodes[nId + next];
threshold = rescale<isUp>(level, node);
sum = get<isUp>(x, y + (node.threshold >> 28) * hogluvStep, node.rect);
const int lShift = (next - 1) * 2 + (int)(sum >= threshold);
float impact = leaves[(st + threadIdx.x) * 4 + lShift];
PrefixSum<Policy>::apply(impact);
confidence += impact;
if(__any((confidence <= stages[(st + threadIdx.x)]))) st += 2048;
}
if(!threadIdx.x && st == stEnd && ((confidence - FLT_EPSILON) >= 0))
{
int idx = atomicInc(ctr, ndetections);
objects[idx] = Detection(__float2int_rn(x * Policy::SHRINKAGE),
__float2int_rn(y * Policy::SHRINKAGE), level.objSize.x, level.objSize.y, confidence);
}
}
template<typename Policy, bool isUp>
__global__ void soft_cascade(const CascadeInvoker<Policy> invoker, Detection* objects, const uint n, uint* ctr, const int downs)
{
invoker.template detect<isUp>(objects, n, ctr, downs);
}
template<typename Policy>
void CascadeInvoker<Policy>::operator()(const PtrStepSzb& roi, const PtrStepSzi& hogluv,
PtrStepSz<uchar4> objects, const int downscales, const cudaStream_t& stream) const
{
int fw = roi.rows;
int fh = roi.cols;
dim3 grid(fw, fh / Policy::STA_Y, downscales);
uint* ctr = (uint*)(objects.ptr(0));
Detection* det = ((Detection*)objects.ptr(0)) + 1;
uint max_det = objects.cols / sizeof(Detection);
cudaChannelFormatDesc desc = cudaCreateChannelDesc<int>();
cudaSafeCall( cudaBindTexture2D(0, thogluv, hogluv.data, desc, hogluv.cols, hogluv.rows, hogluv.step));
cudaChannelFormatDesc desc_roi = cudaCreateChannelDesc<typename Policy::roi_type>();
cudaSafeCall( cudaBindTexture2D(0, troi, roi.data, desc_roi, roi.cols / Policy::STA_Y, roi.rows, roi.step));
const CascadeInvoker<Policy> inv = *this;
soft_cascade<Policy, false><<<grid, Policy::block(), 0, stream>>>(inv, det, max_det, ctr, 0);
cudaSafeCall( cudaGetLastError());
grid = dim3(fw, fh / Policy::STA_Y, scales - downscales);
soft_cascade<Policy, true><<<grid, Policy::block(), 0, stream>>>(inv, det, max_det, ctr, downscales);
if (!stream)
{
cudaSafeCall( cudaGetLastError());
cudaSafeCall( cudaDeviceSynchronize());
}
}
template void CascadeInvoker<GK107PolicyX4>::operator()(const PtrStepSzb& roi, const PtrStepSzi& hogluv,
PtrStepSz<uchar4> objects, const int downscales, const cudaStream_t& stream) const;
}
}}}

82
modules/gpu/src/cuda/integral_image.cu
View File

@@ -383,6 +383,88 @@ namespace cv { namespace gpu { namespace device
if (stream == 0)
cudaSafeCall( cudaDeviceSynchronize() );
}
__global__ void shfl_integral_vertical(PtrStepSz<unsigned int> buffer, PtrStepSz<unsigned int> integral)
{
#if defined(__CUDA_ARCH__) && (__CUDA_ARCH__ >= 300)
__shared__ unsigned int sums[32][9];
const int tidx = blockIdx.x * blockDim.x + threadIdx.x;
const int lane_id = tidx % 8;
if (tidx >= integral.cols)
return;
sums[threadIdx.x][threadIdx.y] = 0;
__syncthreads();
unsigned int stepSum = 0;
for (int y = threadIdx.y; y < integral.rows; y += blockDim.y)
{
unsigned int* p = buffer.ptr(y) + tidx;
unsigned int* dst = integral.ptr(y + 1) + tidx + 1;
unsigned int sum = *p;
sums[threadIdx.x][threadIdx.y] = sum;
__syncthreads();
// place into SMEM
// shfl scan reduce the SMEM, reformating so the column
// sums are computed in a warp
// then read out properly
const int j = threadIdx.x % 8;
const int k = threadIdx.x / 8 + threadIdx.y * 4;
int partial_sum = sums[k][j];
for (int i = 1; i <= 8; i *= 2)
{
int n = __shfl_up(partial_sum, i, 32);
if (lane_id >= i)
partial_sum += n;
}
sums[k][j] = partial_sum;
__syncthreads();
if (threadIdx.y > 0)
sum += sums[threadIdx.x][threadIdx.y - 1];
sum += stepSum;
stepSum += sums[threadIdx.x][blockDim.y - 1];
__syncthreads();
*dst = sum;
}
#endif
}
// used for frame preprocessing before Soft Cascade evaluation: no synchronization needed
void shfl_integral_gpu_buffered(PtrStepSzb img, PtrStepSz<uint4> buffer, PtrStepSz<unsigned int> integral,
int blockStep, cudaStream_t stream)
{
{
const int block = blockStep;
const int grid = img.rows;
cudaSafeCall( cudaFuncSetCacheConfig(shfl_integral_horizontal, cudaFuncCachePreferL1) );
shfl_integral_horizontal<<<grid, block, 0, stream>>>((PtrStepSz<uint4>) img, buffer);
cudaSafeCall( cudaGetLastError() );
}
{
const dim3 block(32, 8);
const dim3 grid(divUp(integral.cols, block.x), 1);
shfl_integral_vertical<<<grid, block, 0, stream>>>((PtrStepSz<uint>)buffer, integral);
cudaSafeCall( cudaGetLastError() );
}
}
}
}}}

2
modules/gpu/src/cuda/texture_binder.hpp
View File

@@ -85,7 +85,7 @@ namespace cv
namespace device
{
using pcl::gpu::TextureBinder;
using cv::gpu::TextureBinder;
}
}

60
modules/gpu/src/gpu_init.cpp Normal file
View File

@@ -0,0 +1,60 @@
/*M///////////////////////////////////////////////////////////////////////////////////////
//
// IMPORTANT: READ BEFORE DOWNLOADING, COPYING, INSTALLING OR USING.
//
// By downloading, copying, installing or using the software you agree to this license.
// If you do not agree to this license, do not download, install,
// copy or use the software.
//
//
// License Agreement
// For Open Source Computer Vision Library
//
// Copyright (C) 2000-2008, Intel Corporation, all rights reserved.
// Copyright (C) 2008-2012, Willow Garage Inc., all rights reserved.
// Third party copyrights are property of their respective owners.
//
// Redistribution and use in source and binary forms, with or without modification,
// are permitted provided that the following conditions are met:
//
// * Redistribution's of source code must retain the above copyright notice,
// this list of conditions and the following disclaimer.
//
// * Redistribution's in binary form must reproduce the above copyright notice,
// this list of conditions and the following disclaimer in the documentation
// and/or other materials provided with the distribution.
//
// * The name of the copyright holders may not be used to endorse or promote products
// derived from this software without specific prior written permission.
//
// This software is provided by the copyright holders and contributors "as is" and
// any express or implied warranties, including, but not limited to, the implied
// warranties of merchantability and fitness for a particular purpose are disclaimed.
// In no event shall the Intel Corporation or contributors be liable for any direct,
// indirect, incidental, special, exemplary, or consequential damages
// (including, but not limited to, procurement of substitute goods or services;
// loss of use, data, or profits; or business interruption) however caused
// and on any theory of liability, whether in contract, strict liability,
// or tort (including negligence or otherwise) arising in any way out of
// the use of this software, even if advised of the possibility of such damage.
//
//M*/
#include <precomp.hpp>
namespace cv { namespace gpu
{
CV_INIT_ALGORITHM(SCascade, "CascadeDetector.SCascade",
obj.info()->addParam(obj, "minScale", obj.minScale);
obj.info()->addParam(obj, "maxScale", obj.maxScale);
obj.info()->addParam(obj, "scales", obj.scales);
obj.info()->addParam(obj, "rejCriteria", obj.rejCriteria));
bool initModule_gpu(void)
{
Ptr<Algorithm> sc = createSCascade();
return sc->info() != 0;
}
} }

153
modules/gpu/src/icf.hpp Normal file
View File

@@ -0,0 +1,153 @@
//M///////////////////////////////////////////////////////////////////////////////////////
//
// IMPORTANT: READ BEFORE DOWNLOADING, COPYING, INSTALLING OR USING.
//
// By downloading, copying, installing or using the software you agree to this license.
// If you do not agree to this license, do not download, install,
// copy or use the software.
//
//
// License Agreement
// For Open Source Computer Vision Library
//
// Copyright (C) 2000-2008, Intel Corporation, all rights reserved.
// Copyright (C) 2008-2012, Willow Garage Inc., all rights reserved.
// Third party copyrights are property of their respective owners.
//
// Redistribution and use in source and binary forms, with or without modification,
// are permitted provided that the following conditions are met:
//
// * Redistribution's of source code must retain the above copyright notice,
// this list of conditions and the following disclaimer.
//
// * Redistribution's in binary form must reproduce the above copyright notice,
// this list of conditions and the following disclaimer in the documentation
// and/or other materials provided with the distribution.
//
// * The name of the copyright holders may not be used to endorse or promote products
// derived from this software without specific prior written permission.
//
// This software is provided by the copyright holders and contributors "as is" and
// any express or implied warranties, including, but not limited to, the implied
// warranties of merchantability and fitness for a particular purpose are disclaimed.
// In no event shall the Intel Corporation or contributors be liable for any direct,
// indirect, incidental, special, exemplary, or consequential damages
// (including, but not limited to, procurement of substitute goods or services;
// loss of use, data, or profits; or business interruption) however caused
// and on any theory of liability, whether in contract, strict liability,
// or tort (including negligence or otherwise) arising in any way out of
// the use of this software, even if advised of the possibility of such damage.
//
//M
#ifndef __OPENCV_ICF_HPP__
#define __OPENCV_ICF_HPP__
#include <opencv2/gpu/device/common.hpp>
#if defined __CUDACC__
# define __device __device__ __forceinline__
#else
# define __device
#endif
namespace cv { namespace gpu { namespace device {
namespace icf {
struct __align__(16) Octave
{
ushort index;
ushort stages;
ushort shrinkage;
ushort2 size;
float scale;
Octave(const ushort i, const ushort s, const ushort sh, const ushort2 sz, const float sc)
: index(i), stages(s), shrinkage(sh), size(sz), scale(sc) {}
};
struct __align__(8) Level //is actually 24 bytes
{
int octave;
int step;
float relScale;
float scaling[2]; // calculated according to Dollal paper
// for 640x480 we can not get overflow
uchar2 workRect;
uchar2 objSize;
Level(int idx, const Octave& oct, const float scale, const int w, const int h);
__device Level(){}
};
struct __align__(8) Node
{
uchar4 rect;
// ushort channel;
uint threshold;
enum { THRESHOLD_MASK = 0x0FFFFFFF };
Node(const uchar4 r, const uint ch, const uint t) : rect(r), threshold(t + (ch << 28)) {}
};
struct __align__(16) Detection
{
ushort x;
ushort y;
ushort w;
ushort h;
float confidence;
int kind;
Detection(){}
__device Detection(int _x, int _y, uchar _w, uchar _h, float c)
: x(_x), y(_y), w(_w), h(_h), confidence(c), kind(0) {};
};
struct GK107PolicyX4
{
enum {WARP = 32, STA_X = WARP, STA_Y = 8, SHRINKAGE = 4};
typedef float2 roi_type;
static const dim3 block()
{
return dim3(STA_X, STA_Y);
}
};
template<typename Policy>
struct CascadeInvoker
{
CascadeInvoker(): levels(0), stages(0), nodes(0), leaves(0), scales(0) {}
CascadeInvoker(const PtrStepSzb& _levels, const PtrStepSzf& _stages,
const PtrStepSzb& _nodes, const PtrStepSzf& _leaves)
: levels((const Level*)_levels.ptr()),
stages((const float*)_stages.ptr()),
nodes((const Node*)_nodes.ptr()), leaves((const float*)_leaves.ptr()),
scales(_levels.cols / sizeof(Level))
{}
const Level* levels;
const float* stages;
const Node* nodes;
const float* leaves;
int scales;
void operator()(const PtrStepSzb& roi, const PtrStepSzi& hogluv, PtrStepSz<uchar4> objects,
const int downscales, const cudaStream_t& stream = 0) const;
template<bool isUp>
__device void detect(Detection* objects, const uint ndetections, uint* ctr, const int downscales) const;
};
}
}}}
#endif

2
modules/gpu/src/nvidia/core/NCV.hpp
View File

@@ -288,7 +288,7 @@ NCV_EXPORTS void ncvSetDebugOutputHandler(NCVDebugOutputHandler* func);
do \
{ \
cudaError_t res = cudacall; \
ncvAssertPrintReturn(cudaSuccess==res, "cudaError_t=" << res, errCode); \
ncvAssertPrintReturn(cudaSuccess==res, "cudaError_t=" << (int)res, errCode); \
} while (0)

603
modules/gpu/src/softcascade.cpp Normal file
View File

@@ -0,0 +1,603 @@
/*M///////////////////////////////////////////////////////////////////////////////////////
//
// IMPORTANT: READ BEFORE DOWNLOADING, COPYING, INSTALLING OR USING.
//
// By downloading, copying, installing or using the software you agree to this license.
// If you do not agree to this license, do not download, install,
// copy or use the software.
//
//
// License Agreement
// For Open Source Computer Vision Library
//
// Copyright (C) 2000-2008, Intel Corporation, all rights reserved.
// Copyright (C) 2008-2012, Willow Garage Inc., all rights reserved.
// Third party copyrights are property of their respective owners.
//
// Redistribution and use in source and binary forms, with or without modification,
// are permitted provided that the following conditions are met:
//
// * Redistribution's of source code must retain the above copyright notice,
// this list of conditions and the following disclaimer.
//
// * Redistribution's in binary form must reproduce the above copyright notice,
// this list of conditions and the following disclaimer in the documentation
// and/or other materials provided with the distribution.
//
// * The name of the copyright holders may not be used to endorse or promote products
// derived from this software without specific prior written permission.
//
// This software is provided by the copyright holders and contributors "as is" and
// any express or implied warranties, including, but not limited to, the implied
// warranties of merchantability and fitness for a particular purpose are disclaimed.
// In no event shall the Intel Corporation or contributors be liable for any direct,
// indirect, incidental, special, exemplary, or consequential damages
// (including, but not limited to, procurement of substitute goods or services;
// loss of use, data, or profits; or business interruption) however caused
// and on any theory of liability, whether in contract, strict liability,
// or tort (including negligence or otherwise) arising in any way out of
// the use of this software, even if advised of the possibility of such damage.
//
//M*/
#include <precomp.hpp>
#include <opencv2/highgui/highgui.hpp>
#if !defined (HAVE_CUDA)
cv::gpu::SCascade::SCascade(const double, const double, const int, const int) { throw_nogpu(); }
cv::gpu::SCascade::~SCascade() { throw_nogpu(); }
bool cv::gpu::SCascade::load(const FileNode&) { throw_nogpu(); return false;}
void cv::gpu::SCascade::detect(InputArray, InputArray, OutputArray, Stream&) const { throw_nogpu(); }
void cv::gpu::SCascade::genRoi(InputArray, OutputArray, Stream&) const { throw_nogpu(); }
void cv::gpu::SCascade::read(const FileNode& fn) { Algorithm::read(fn); }
#else
#include <icf.hpp>
cv::gpu::device::icf::Level::Level(int idx, const Octave& oct, const float scale, const int w, const int h)
: octave(idx), step(oct.stages), relScale(scale / oct.scale)
{
workRect.x = round(w / (float)oct.shrinkage);
workRect.y = round(h / (float)oct.shrinkage);
objSize.x = cv::saturate_cast<uchar>(oct.size.x * relScale);
objSize.y = cv::saturate_cast<uchar>(oct.size.y * relScale);
// according to R. Benenson, M. Mathias, R. Timofte and L. Van Gool's and Dallal's papers
if (fabs(relScale - 1.f) < FLT_EPSILON)
scaling[0] = scaling[1] = 1.f;
else
{
scaling[0] = (relScale < 1.f) ? 0.89f * ::pow(relScale, 1.099f / ::log(2)) : 1.f;
scaling[1] = relScale * relScale;
}
}
namespace cv { namespace gpu { namespace device {
namespace icf {
void fillBins(cv::gpu::PtrStepSzb hogluv, const cv::gpu::PtrStepSzf& nangle,
const int fw, const int fh, const int bins, cudaStream_t stream);
void suppress(const PtrStepSzb& objects, PtrStepSzb overlaps, PtrStepSzi ndetections,
PtrStepSzb suppressed, cudaStream_t stream);
}
namespace imgproc {
void shfl_integral_gpu_buffered(PtrStepSzb, PtrStepSz<uint4>, PtrStepSz<unsigned int>, int, cudaStream_t);
template <typename T>
void resize_gpu(PtrStepSzb src, PtrStepSzb srcWhole, int xoff, int yoff, float fx, float fy,
PtrStepSzb dst, int interpolation, cudaStream_t stream);
}
}}}
struct cv::gpu::SCascade::Fields
{
static Fields* parseCascade(const FileNode &root, const float mins, const float maxs, const int totals)
{
static const char *const SC_STAGE_TYPE = "stageType";
static const char *const SC_BOOST = "BOOST";
static const char *const SC_FEATURE_TYPE = "featureType";
static const char *const SC_ICF = "ICF";
// only Ada Boost supported
std::string stageTypeStr = (string)root[SC_STAGE_TYPE];
CV_Assert(stageTypeStr == SC_BOOST);
// only HOG-like integral channel features cupported
string featureTypeStr = (string)root[SC_FEATURE_TYPE];
CV_Assert(featureTypeStr == SC_ICF);
static const char *const SC_ORIG_W = "width";
static const char *const SC_ORIG_H = "height";
int origWidth = (int)root[SC_ORIG_W];
int origHeight = (int)root[SC_ORIG_H];
static const char *const SC_OCTAVES = "octaves";
static const char *const SC_STAGES = "stages";
static const char *const SC_FEATURES = "features";
static const char *const SC_WEEK = "weakClassifiers";
static const char *const SC_INTERNAL = "internalNodes";
static const char *const SC_LEAF = "leafValues";
static const char *const SC_OCT_SCALE = "scale";
static const char *const SC_OCT_STAGES = "stageNum";
static const char *const SC_OCT_SHRINKAGE = "shrinkingFactor";
static const char *const SC_STAGE_THRESHOLD = "stageThreshold";
static const char * const SC_F_CHANNEL = "channel";
static const char * const SC_F_RECT = "rect";
FileNode fn = root[SC_OCTAVES];
if (fn.empty()) return false;
using namespace device::icf;
std::vector<Octave> voctaves;
std::vector<float> vstages;
std::vector<Node> vnodes;
std::vector<float> vleaves;
FileNodeIterator it = fn.begin(), it_end = fn.end();
int feature_offset = 0;
ushort octIndex = 0;
ushort shrinkage = 1;
for (; it != it_end; ++it)
{
FileNode fns = *it;
float scale = (float)fns[SC_OCT_SCALE];
bool isUPOctave = scale >= 1;
ushort nstages = saturate_cast<ushort>((int)fns[SC_OCT_STAGES]);
ushort2 size;
size.x = cvRound(origWidth * scale);
size.y = cvRound(origHeight * scale);
shrinkage = saturate_cast<ushort>((int)fns[SC_OCT_SHRINKAGE]);
Octave octave(octIndex, nstages, shrinkage, size, scale);
CV_Assert(octave.stages > 0);
voctaves.push_back(octave);
FileNode ffs = fns[SC_FEATURES];
if (ffs.empty()) return false;
FileNodeIterator ftrs = ffs.begin();
fns = fns[SC_STAGES];
if (fn.empty()) return false;
// for each stage (~ decision tree with H = 2)
FileNodeIterator st = fns.begin(), st_end = fns.end();
for (; st != st_end; ++st )
{
fns = *st;
vstages.push_back((float)fns[SC_STAGE_THRESHOLD]);
fns = fns[SC_WEEK];
FileNodeIterator ftr = fns.begin(), ft_end = fns.end();
for (; ftr != ft_end; ++ftr)
{
fns = (*ftr)[SC_INTERNAL];
FileNodeIterator inIt = fns.begin(), inIt_end = fns.end();
for (; inIt != inIt_end;)
{
// int feature = (int)(*(inIt +=2)) + feature_offset;
inIt +=3;
// extract feature, Todo:check it
uint th = saturate_cast<uint>((float)(*(inIt++)));
cv::FileNode ftn = (*ftrs)[SC_F_RECT];
cv::FileNodeIterator r_it = ftn.begin();
uchar4 rect;
rect.x = saturate_cast<uchar>((int)*(r_it++));
rect.y = saturate_cast<uchar>((int)*(r_it++));
rect.z = saturate_cast<uchar>((int)*(r_it++));
rect.w = saturate_cast<uchar>((int)*(r_it++));
if (isUPOctave)
{
rect.z -= rect.x;
rect.w -= rect.y;
}
uint channel = saturate_cast<uint>((int)(*ftrs)[SC_F_CHANNEL]);
vnodes.push_back(Node(rect, channel, th));
++ftrs;
}
fns = (*ftr)[SC_LEAF];
inIt = fns.begin(), inIt_end = fns.end();
for (; inIt != inIt_end; ++inIt)
vleaves.push_back((float)(*inIt));
}
}
feature_offset += octave.stages * 3;
++octIndex;
}
cv::Mat hoctaves(1, voctaves.size() * sizeof(Octave), CV_8UC1, (uchar*)&(voctaves[0]));
CV_Assert(!hoctaves.empty());
cv::Mat hstages(cv::Mat(vstages).reshape(1,1));
CV_Assert(!hstages.empty());
cv::Mat hnodes(1, vnodes.size() * sizeof(Node), CV_8UC1, (uchar*)&(vnodes[0]) );
CV_Assert(!hnodes.empty());
cv::Mat hleaves(cv::Mat(vleaves).reshape(1,1));
CV_Assert(!hleaves.empty());
Fields* fields = new Fields(mins, maxs, totals, origWidth, origHeight, shrinkage, 0,
hoctaves, hstages, hnodes, hleaves);
fields->voctaves = voctaves;
fields->createLevels(FRAME_HEIGHT, FRAME_WIDTH);
return fields;
}
bool check(float mins,float maxs, int scales)
{
bool updated = (minScale == mins) || (maxScale == maxs) || (totals = scales);
minScale = mins;
maxScale = maxScale;
totals = scales;
return updated;
}
int createLevels(const int fh, const int fw)
{
using namespace device::icf;
std::vector<Level> vlevels;
float logFactor = (::log(maxScale) - ::log(minScale)) / (totals -1);
float scale = minScale;
int dcs = 0;
for (int sc = 0; sc < totals; ++sc)
{
int width = ::std::max(0.0f, fw - (origObjWidth * scale));
int height = ::std::max(0.0f, fh - (origObjHeight * scale));
float logScale = ::log(scale);
int fit = fitOctave(voctaves, logScale);
Level level(fit, voctaves[fit], scale, width, height);
if (!width || !height)
break;
else
{
vlevels.push_back(level);
if (voctaves[fit].scale < 1) ++dcs;
}
if (::fabs(scale - maxScale) < FLT_EPSILON) break;
scale = ::std::min(maxScale, ::expf(::log(scale) + logFactor));
}
cv::Mat hlevels = cv::Mat(1, vlevels.size() * sizeof(Level), CV_8UC1, (uchar*)&(vlevels[0]) );
CV_Assert(!hlevels.empty());
levels.upload(hlevels);
downscales = dcs;
return dcs;
}
bool update(int fh, int fw, int shr)
{
if ((fh == luv.rows) && (fw == luv.cols)) return false;
plane.create(fh * (HOG_LUV_BINS + 1), fw, CV_8UC1);
fplane.create(fh * HOG_BINS, fw, CV_32FC1);
luv.create(fh, fw, CV_8UC3);
shrunk.create(fh / shr * HOG_LUV_BINS, fw / shr, CV_8UC1);
integralBuffer.create(shrunk.rows, shrunk.cols, CV_32SC1);
hogluv.create((fh / shr) * HOG_LUV_BINS + 1, fw / shr + 1, CV_32SC1);
hogluv.setTo(cv::Scalar::all(0));
overlaps.create(1, 5000, CV_8UC1);
suppressed.create(1, sizeof(Detection) * 51, CV_8UC1);
return true;
}
Fields( const float mins, const float maxs, const int tts, const int ow, const int oh, const int shr, const int ds,
cv::Mat hoctaves, cv::Mat hstages, cv::Mat hnodes, cv::Mat hleaves)
: minScale(mins), maxScale(maxs), totals(tts), origObjWidth(ow), origObjHeight(oh), shrinkage(shr), downscales(ds)
{
update(FRAME_HEIGHT, FRAME_WIDTH, shr);
octaves.upload(hoctaves);
stages.upload(hstages);
nodes.upload(hnodes);
leaves.upload(hleaves);
}
void detect(const cv::gpu::GpuMat& roi, cv::gpu::GpuMat& objects, Stream& s) const
{
if (s)
s.enqueueMemSet(objects, 0);
else
cudaMemset(objects.data, 0, sizeof(Detection));
cudaSafeCall( cudaGetLastError());
device::icf::CascadeInvoker<device::icf::GK107PolicyX4> invoker
= device::icf::CascadeInvoker<device::icf::GK107PolicyX4>(levels, stages, nodes, leaves);
cudaStream_t stream = StreamAccessor::getStream(s);
invoker(roi, hogluv, objects, downscales, stream);
}
void preprocess(const cv::gpu::GpuMat& colored, Stream& s)
{
if (s)
s.enqueueMemSet(plane, 0);
else
cudaMemset(plane.data, 0, plane.step * plane.rows);
const int fw = colored.cols;
const int fh = colored.rows;
GpuMat gray(plane, cv::Rect(0, fh * Fields::HOG_LUV_BINS, fw, fh));
cv::gpu::cvtColor(colored, gray, CV_BGR2GRAY, s);
createHogBins(gray ,s);
createLuvBins(colored, s);
integrate(fh, fw, s);
}
void suppress(GpuMat& objects, Stream& s)
{
GpuMat ndetections = GpuMat(objects, cv::Rect(0, 0, sizeof(Detection), 1));
ensureSizeIsEnough(objects.rows, objects.cols, CV_8UC1, overlaps);
if (s)
{
s.enqueueMemSet(overlaps, 0);
s.enqueueMemSet(suppressed, 0);
}
else
{
overlaps.setTo(0);
suppressed.setTo(0);
}
cudaStream_t stream = StreamAccessor::getStream(s);
device::icf::suppress(objects, overlaps, ndetections, suppressed, stream);
}
private:
typedef std::vector<device::icf::Octave>::const_iterator octIt_t;
static int fitOctave(const std::vector<device::icf::Octave>& octs, const float& logFactor)
{
float minAbsLog = FLT_MAX;
int res = 0;
for (int oct = 0; oct < (int)octs.size(); ++oct)
{
const device::icf::Octave& octave =octs[oct];
float logOctave = ::log(octave.scale);
float logAbsScale = ::fabs(logFactor - logOctave);
if(logAbsScale < minAbsLog)
{
res = oct;
minAbsLog = logAbsScale;
}
}
return res;
}
void createHogBins(const cv::gpu::GpuMat& gray, Stream& s)
{
static const int fw = gray.cols;
static const int fh = gray.rows;
GpuMat dfdx(fplane, cv::Rect(0, 0, fw, fh));
GpuMat dfdy(fplane, cv::Rect(0, fh, fw, fh));
cv::gpu::Sobel(gray, dfdx, CV_32F, 1, 0, sobelBuf, 3, 1, BORDER_DEFAULT, -1, s);
cv::gpu::Sobel(gray, dfdy, CV_32F, 0, 1, sobelBuf, 3, 1, BORDER_DEFAULT, -1, s);
GpuMat mag(fplane, cv::Rect(0, 2 * fh, fw, fh));
GpuMat ang(fplane, cv::Rect(0, 3 * fh, fw, fh));
cv::gpu::cartToPolar(dfdx, dfdy, mag, ang, true, s);
// normolize magnitude to uchar interval and angles to 6 bins
GpuMat nmag(fplane, cv::Rect(0, 4 * fh, fw, fh));
GpuMat nang(fplane, cv::Rect(0, 5 * fh, fw, fh));
cv::gpu::multiply(mag, cv::Scalar::all(1.f / (8 *::log(2))), nmag, 1, -1, s);
cv::gpu::multiply(ang, cv::Scalar::all(1.f / 60.f), nang, 1, -1, s);
//create uchar magnitude
GpuMat cmag(plane, cv::Rect(0, fh * Fields::HOG_BINS, fw, fh));
if (s)
s.enqueueConvert(nmag, cmag, CV_8UC1);
else
nmag.convertTo(cmag, CV_8UC1);
cudaStream_t stream = StreamAccessor::getStream(s);
device::icf::fillBins(plane, nang, fw, fh, Fields::HOG_BINS, stream);
}
void createLuvBins(const cv::gpu::GpuMat& colored, Stream& s)
{
static const int fw = colored.cols;
static const int fh = colored.rows;
cv::gpu::cvtColor(colored, luv, CV_BGR2Luv, s);
std::vector<GpuMat> splited;
for(int i = 0; i < Fields::LUV_BINS; ++i)
{
splited.push_back(GpuMat(plane, cv::Rect(0, fh * (7 + i), fw, fh)));
}
cv::gpu::split(luv, splited, s);
}
void integrate(const int fh, const int fw, Stream& s)
{
GpuMat channels(plane, cv::Rect(0, 0, fw, fh * Fields::HOG_LUV_BINS));
cv::gpu::resize(channels, shrunk, cv::Size(), 1.f / shrinkage, 1.f / shrinkage, CV_INTER_AREA, s);
if (info.majorVersion() < 3)
cv::gpu::integralBuffered(shrunk, hogluv, integralBuffer, s);
else
{
cudaStream_t stream = StreamAccessor::getStream(s);
device::imgproc::shfl_integral_gpu_buffered(shrunk, integralBuffer, hogluv, 12, stream);
}
}
public:
// scales range
float minScale;
float maxScale;
int totals;
int origObjWidth;
int origObjHeight;
const int shrinkage;
int downscales;
// preallocated buffer 640x480x10 for hogluv + 640x480 got gray
GpuMat plane;
// preallocated buffer for floating point operations
GpuMat fplane;
// temporial mat for cvtColor
GpuMat luv;
// 160x120x10
GpuMat shrunk;
// temporial mat for integrall
GpuMat integralBuffer;
// 161x121x10
GpuMat hogluv;
// used for area overlap computing during
GpuMat overlaps;
// used for suppression
GpuMat suppressed;
// Cascade from xml
GpuMat octaves;
GpuMat stages;
GpuMat nodes;
GpuMat leaves;
GpuMat levels;
GpuMat sobelBuf;
GpuMat collected;
std::vector<device::icf::Octave> voctaves;
DeviceInfo info;
enum { BOOST = 0 };
enum
{
FRAME_WIDTH = 640,
FRAME_HEIGHT = 480,
HOG_BINS = 6,
LUV_BINS = 3,
HOG_LUV_BINS = 10
};
};
cv::gpu::SCascade::SCascade(const double mins, const double maxs, const int sc, const int rjf)
: fields(0), minScale(mins), maxScale(maxs), scales(sc), rejCriteria(rjf) {}
cv::gpu::SCascade::~SCascade() { delete fields; }
bool cv::gpu::SCascade::load(const FileNode& fn)
{
if (fields) delete fields;
fields = Fields::parseCascade(fn, minScale, maxScale, scales);
return fields != 0;
}
void cv::gpu::SCascade::detect(InputArray image, InputArray _rois, OutputArray _objects, Stream& s) const
{
CV_Assert(fields);
const GpuMat colored = image.getGpuMat();
// only color images are supperted
CV_Assert(colored.type() == CV_8UC3 || colored.type() == CV_32SC1);
GpuMat rois = _rois.getGpuMat(), objects = _objects.getGpuMat();
Fields& flds = *fields;
if (colored.type() == CV_8UC3)
{
if (!flds.update(colored.rows, colored.cols, flds.shrinkage) || flds.check(minScale, maxScale, scales))
flds.createLevels(colored.rows, colored.cols);
flds.preprocess(colored, s);
}
else
{
if (s)
s.enqueueCopy(colored, flds.hogluv);
else
colored.copyTo(flds.hogluv);
}
flds.detect(rois, objects, s);
if (rejCriteria != NO_REJECT)
{
GpuMat spr(objects, cv::Rect(0, 0, flds.suppressed.cols, flds.suppressed.rows));
flds.suppress(objects, s);
flds.suppressed.copyTo(spr);
}
}
void cv::gpu::SCascade::genRoi(InputArray _roi, OutputArray _mask, Stream& stream) const
{
CV_Assert(fields);
int shr = (*fields).shrinkage;
const GpuMat roi = _roi.getGpuMat();
_mask.create( roi.cols / shr, roi.rows / shr, roi.type() );
GpuMat mask = _mask.getGpuMat();
cv::gpu::GpuMat tmp;
cv::gpu::resize(roi, tmp, cv::Size(), 1.f / shr, 1.f / shr, CV_INTER_AREA, stream);
cv::gpu::transpose(tmp, mask, stream);
}
void cv::gpu::SCascade::read(const FileNode& fn)
{
Algorithm::read(fn);
}
#endif