Added getLines, fill_accum_local kernels

This commit is contained in:
Alexander Karsakov 2014-08-25 15:57:58 +04:00
parent 038bfb98ec
commit f7aadd07f6
4 changed files with 169 additions and 44 deletions

View File

@ -239,8 +239,9 @@ namespace
void GeneralizedHoughBase::detectImpl(InputArray image, OutputArray positions, OutputArray votes)
{
#ifndef HAVE_OPENCV_CUDAFILTERS
(void) templ;
(void) templCenter;
(void) image;
(void) positions;
(void) votes;
throw_no_cuda();
#else
calcEdges(image, imageEdges_, imageDx_, imageDy_);

View File

@ -668,9 +668,10 @@ static bool ocl_HoughLines(InputArray _src, OutputArray _lines, double rho, doub
UMat src = _src.getUMat();
float irho = 1 / rho;
float irho = (float) (1 / rho);
int numangle = cvRound((max_theta - min_theta) / theta);
int numrho = cvRound(((src.cols + src.rows) * 2 + 1) / rho);
ocl::Device dev = ocl::Device::getDefault();
// make list of nonzero points
const int pixelsPerWI = 4;
@ -680,7 +681,7 @@ static bool ocl_HoughLines(InputArray _src, OutputArray _lines, double rho, doub
if (pointListKernel.empty())
return false;
UMat pointsList(1, src.total(), CV_32SC1);
UMat pointsList(1, (int) src.total(), CV_32SC1);
UMat total(1, 1, CV_32SC1, Scalar::all(0));
pointListKernel.args(ocl::KernelArg::ReadOnly(src), ocl::KernelArg::WriteOnlyNoSize(pointsList),
ocl::KernelArg::PtrWriteOnly(total));
@ -692,37 +693,66 @@ static bool ocl_HoughLines(InputArray _src, OutputArray _lines, double rho, doub
int total_points = total.getMat(ACCESS_READ).at<int>(0, 0);
if (total_points <= 0)
return false;
{
_lines.assign(UMat(0,0,CV_32FC2));
return true;
}
// convert src to hough space
group_size = (total_points + pixelsPerWI - 1)/pixelsPerWI;
ocl::Kernel fillAccumKernel("fill_accum", ocl::imgproc::hough_lines_oclsrc,
format("-D FILL_ACCUM -D GROUP_SIZE=%d", group_size));
group_size = min((int) dev.maxWorkGroupSize(), total_points);
int local_memory_needed = (numrho + 2)*sizeof(int);
ocl::Kernel fillAccumKernel;
globalThreads[0] = group_size; globalThreads[1] = numangle;
size_t* fillAccumLT = NULL;
UMat accum(numangle + 2, numrho + 2, CV_32SC1);
if (local_memory_needed > dev.localMemSize())
{
fillAccumKernel.create("fill_accum_global", ocl::imgproc::hough_lines_oclsrc,
format("-D FILL_ACCUM_GLOBAL"));
accum.setTo(Scalar::all(0));
}
else
{
fillAccumKernel.create("fill_accum_local", ocl::imgproc::hough_lines_oclsrc,
format("-D FILL_ACCUM_LOCAL -D LOCAL_SIZE=%d -D BUFFER_SIZE=%d", group_size, numrho + 2));
localThreads[0] = group_size; localThreads[1] = 1;
fillAccumLT = localThreads;
}
if (fillAccumKernel.empty())
return false;
UMat accum(numangle + 2, numrho + 2, CV_32SC1, Scalar::all(0));
int linesMax = min(total_points*numangle/threshold, 4096);
UMat lines(linesMax, 1, CV_32FC2);
UMat lines_count(1, 1, CV_32SC1, Scalar::all(0));
fillAccumKernel.args(ocl::KernelArg::ReadOnlyNoSize(pointsList), ocl::KernelArg::WriteOnly(accum),
total_points, irho, (float) theta, numrho, numangle);
globalThreads[0] = group_size; globalThreads[1] = numangle;
if (!fillAccumKernel.run(2, globalThreads, NULL, false))
return false;
printf("GPU: \n");
int sum = 0;
Mat ac = accum.getMat(ACCESS_READ);
for (int i=0; i<8; i++)
{
for (int j=0; j<8; j++)
{
sum += ac.at<int>(i, j);
printf("%d ", ac.at<int>(i, j));
}
printf("\n");
}
printf("sum = %d\n", sum);
if (!fillAccumKernel.run(2, globalThreads, fillAccumLT, false))
return false;
ocl::Kernel getLinesKernel("get_lines", ocl::imgproc::hough_lines_oclsrc,
format("-D GET_LINES"));
if (getLinesKernel.empty())
return false;
globalThreads[0] = numrho; globalThreads[1] = numangle;
getLinesKernel.args(ocl::KernelArg::ReadOnly(accum), ocl::KernelArg::WriteOnlyNoSize(lines),
ocl::KernelArg::PtrWriteOnly(lines_count), linesMax, threshold, (float) rho, (float) theta);
if (!getLinesKernel.run(2, globalThreads, NULL, false))
return false;
int total_lines = min(lines_count.getMat(ACCESS_READ).at<int>(0, 0), linesMax);
if (total_lines > 0)
_lines.assign(lines.rowRange(Range(0, total_lines)));
else
_lines.assign(UMat(0,0,CV_32FC2));
return true;
}
}

View File

@ -50,9 +50,9 @@ __kernel void make_point_list(__global const uchar * src_ptr, int src_step, int
}
}
#elif defined FILL_ACCUM
#elif defined FILL_ACCUM_GLOBAL
__kernel void fill_accum(__global const uchar * list_ptr, int list_step, int list_offset,
__kernel void fill_accum_global(__global const uchar * list_ptr, int list_step, int list_offset,
__global uchar * accum_ptr, int accum_step, int accum_offset, int accum_rows, int accum_cols,
int count, float irho, float theta, int numrho, int numangle)
{
@ -82,5 +82,82 @@ __kernel void fill_accum(__global const uchar * list_ptr, int list_step, int lis
}
}
#elif defined FILL_ACCUM_LOCAL
__kernel void fill_accum_local(__global const uchar * list_ptr, int list_step, int list_offset,
__global uchar * accum_ptr, int accum_step, int accum_offset, int accum_rows, int accum_cols,
int count, float irho, float theta, int numrho, int numangle)
{
int theta_idx = get_global_id(1);
int count_idx = get_local_id(0);
float cosVal;
float sinVal = sincos(theta * ((float)theta_idx), &cosVal);
sinVal *= irho;
cosVal *= irho;
__local int l_accum[BUFFER_SIZE];
for (int i=count_idx; i<BUFFER_SIZE; i+=LOCAL_SIZE)
l_accum[i] = 0;
barrier(CLK_LOCAL_MEM_FENCE);
__global const int * list = (__global const int*)(list_ptr + list_offset);
const int shift = (numrho - 1) / 2;
if (theta_idx < numangle)
{
for (int i = count_idx; i < count; i += LOCAL_SIZE)
{
const int val = list[i];
const int x = (val & 0xFFFF);
const int y = (val >> 16) & 0xFFFF;
int r = convert_int_rte(x * cosVal + y * sinVal) + shift;
atomic_inc(l_accum + r + 1);
}
}
barrier(CLK_LOCAL_MEM_FENCE);
__global int* accum = (__global int*)(accum_ptr + mad24(theta_idx + 1, accum_step, accum_offset));
for (int i=count_idx; i<BUFFER_SIZE; i+=LOCAL_SIZE)
accum[i] = l_accum[i];
}
#elif defined GET_LINES
#define ACCUM(ptr) *((__global int*)(ptr))
__kernel void get_lines(__global uchar * accum_ptr, int accum_step, int accum_offset, int accum_rows, int accum_cols,
__global uchar * lines_ptr, int lines_step, int lines_offset, __global int* lines_index,
int linesMax, int threshold, float rho, float theta)
{
int x = get_global_id(0);
int y = get_global_id(1);
if (x < accum_cols-2 && y < accum_rows-2)
{
__global uchar* accum = accum_ptr + mad24(y+1, accum_step, mad24(x+1, (int) sizeof(int), accum_offset));
__global float2* lines = (__global float2*)(lines_ptr + lines_offset);
int curVote = ACCUM(accum);
if (curVote > threshold && curVote > ACCUM(accum - sizeof(int)) && curVote >= ACCUM(accum + sizeof(int)) &&
curVote > ACCUM(accum - accum_step) && curVote >= ACCUM(accum + accum_step))
{
int index = atomic_inc(lines_index);
if (index < linesMax)
{
float radius = (x - (accum_cols - 3) * 0.5f) * rho;
float angle = y * theta;
lines[index] = (float2)(radius, angle);
}
}
}
}
#endif

View File

@ -13,49 +13,66 @@
namespace cvtest {
namespace ocl {
PARAM_TEST_CASE(HoughLinesTestBase, bool)
struct Vec2fComparator
{
bool operator()(const cv::Vec2f& a, const cv::Vec2f b) const
{
if(a[0] != b[0]) return a[0] < b[0];
else return a[1] < b[1];
}
};
PARAM_TEST_CASE(HoughLinesTestBase, double, double, int)
{
double rhoStep;
double thetaStep;
int threshold;
bool useRoi;
Size src_size;
Mat src, dst;
UMat usrc, udst;
virtual void SetUp()
{
rhoStep = 10;
thetaStep = 0.5;
threshold = 80;
useRoi = false;
rhoStep = GET_PARAM(0);
thetaStep = GET_PARAM(1);
threshold = GET_PARAM(2);
}
virtual void generateTestData()
{
//Mat image = readImage("shared/pic1.png", IMREAD_GRAYSCALE);
src_size = randomSize(500, 1000);
src.create(src_size, CV_8UC1);
src.setTo(Scalar::all(0));
line(src, Point(0, 100), Point(100, 100), Scalar::all(255), 1);
line(src, Point(0, 200), Point(100, 200), Scalar::all(255), 1);
line(src, Point(0, 400), Point(100, 400), Scalar::all(255), 1);
line(src, Point(100, 0), Point(100, 200), Scalar::all(255), 1);
line(src, Point(200, 0), Point(200, 200), Scalar::all(255), 1);
line(src, Point(400, 0), Point(400, 200), Scalar::all(255), 1);
Mat image = randomMat(Size(20, 10), CV_8UC1, 0, 255, false);
cv::threshold(image, src, 127, 255, THRESH_BINARY);
//Canny(image, src, 100, 150, 3);
src.copyTo(usrc);
}
};
typedef HoughLinesTestBase HoughLines;
OCL_TEST_P(HoughLines, RealImage)
OCL_TEST_P(HoughLines, GeneratedImage)
{
for (int j = 0; j < test_loop_times; j++)
{
generateTestData();
//std::cout << src << std::endl;
OCL_OFF(cv::HoughLines(src, dst, rhoStep, thetaStep, threshold));
OCL_ON(cv::HoughLines(usrc, udst, rhoStep, thetaStep, threshold));
OCL_OFF(cv::HoughLines(src, dst, rhoStep, thetaStep, threshold, 0, 0));
OCL_ON(cv::HoughLines(usrc, udst, rhoStep, thetaStep, threshold, 0, 0));
//Near(1e-5);
}
}
OCL_INSTANTIATE_TEST_CASE_P(Imgproc, HoughLines, Values(true, false));
OCL_INSTANTIATE_TEST_CASE_P(Imgproc, HoughLines, Combine(Values(1, 0.5), // rhoStep
Values(CV_PI / 180.0, CV_PI / 360.0), // thetaStep
Values(80, 150))); // threshold
} } // namespace cvtest::ocl