in modules/imgproc/src/filter.cpp [4005:4186]
static bool ocl_filter2D( InputArray _src, OutputArray _dst, int ddepth,
InputArray _kernel, Point anchor,
double delta, int borderType )
{
int type = _src.type(), sdepth = CV_MAT_DEPTH(type), cn = CV_MAT_CN(type);
ddepth = ddepth < 0 ? sdepth : ddepth;
int dtype = CV_MAKE_TYPE(ddepth, cn), wdepth = std::max(std::max(sdepth, ddepth), CV_32F),
wtype = CV_MAKE_TYPE(wdepth, cn);
if (cn > 4)
return false;
Size ksize = _kernel.size();
if (anchor.x < 0)
anchor.x = ksize.width / 2;
if (anchor.y < 0)
anchor.y = ksize.height / 2;
bool isolated = (borderType & BORDER_ISOLATED) != 0;
borderType &= ~BORDER_ISOLATED;
const cv::ocl::Device &device = cv::ocl::Device::getDefault();
bool doubleSupport = device.doubleFPConfig() > 0;
if (wdepth == CV_64F && !doubleSupport)
return false;
const char * const borderMap[] = { "BORDER_CONSTANT", "BORDER_REPLICATE", "BORDER_REFLECT",
"BORDER_WRAP", "BORDER_REFLECT_101" };
cv::Mat kernelMat = _kernel.getMat();
cv::Size sz = _src.size(), wholeSize;
size_t globalsize[2] = { sz.width, sz.height };
size_t localsize_general[2] = {0, 1};
size_t* localsize = NULL;
ocl::Kernel k;
UMat src = _src.getUMat();
if (!isolated)
{
Point ofs;
src.locateROI(wholeSize, ofs);
}
size_t tryWorkItems = device.maxWorkGroupSize();
if (device.isIntel() && 128 < tryWorkItems)
tryWorkItems = 128;
char cvt[2][40];
// For smaller filter kernels, there is a special kernel that is more
// efficient than the general one.
UMat kernalDataUMat;
if (device.isIntel() && (device.type() & ocl::Device::TYPE_GPU) &&
((ksize.width < 5 && ksize.height < 5) ||
(ksize.width == 5 && ksize.height == 5 && cn == 1)))
{
kernelMat = kernelMat.reshape(0, 1);
String kerStr = ocl::kernelToStr(kernelMat, CV_32F);
int h = isolated ? sz.height : wholeSize.height;
int w = isolated ? sz.width : wholeSize.width;
if (w < ksize.width || h < ksize.height)
return false;
// Figure out what vector size to use for loading the pixels.
int pxLoadNumPixels = cn != 1 || sz.width % 4 ? 1 : 4;
int pxLoadVecSize = cn * pxLoadNumPixels;
// Figure out how many pixels per work item to compute in X and Y
// directions. Too many and we run out of registers.
int pxPerWorkItemX = 1;
int pxPerWorkItemY = 1;
if (cn <= 2 && ksize.width <= 4 && ksize.height <= 4)
{
pxPerWorkItemX = sz.width % 8 ? sz.width % 4 ? sz.width % 2 ? 1 : 2 : 4 : 8;
pxPerWorkItemY = sz.height % 2 ? 1 : 2;
}
else if (cn < 4 || (ksize.width <= 4 && ksize.height <= 4))
{
pxPerWorkItemX = sz.width % 2 ? 1 : 2;
pxPerWorkItemY = sz.height % 2 ? 1 : 2;
}
globalsize[0] = sz.width / pxPerWorkItemX;
globalsize[1] = sz.height / pxPerWorkItemY;
// Need some padding in the private array for pixels
int privDataWidth = ROUNDUP(pxPerWorkItemX + ksize.width - 1, pxLoadNumPixels);
// Make the global size a nice round number so the runtime can pick
// from reasonable choices for the workgroup size
const int wgRound = 256;
globalsize[0] = ROUNDUP(globalsize[0], wgRound);
char build_options[1024];
sprintf(build_options, "-D cn=%d "
"-D ANCHOR_X=%d -D ANCHOR_Y=%d -D KERNEL_SIZE_X=%d -D KERNEL_SIZE_Y=%d "
"-D PX_LOAD_VEC_SIZE=%d -D PX_LOAD_NUM_PX=%d "
"-D PX_PER_WI_X=%d -D PX_PER_WI_Y=%d -D PRIV_DATA_WIDTH=%d -D %s -D %s "
"-D PX_LOAD_X_ITERATIONS=%d -D PX_LOAD_Y_ITERATIONS=%d "
"-D srcT=%s -D srcT1=%s -D dstT=%s -D dstT1=%s -D WT=%s -D WT1=%s "
"-D convertToWT=%s -D convertToDstT=%s %s",
cn, anchor.x, anchor.y, ksize.width, ksize.height,
pxLoadVecSize, pxLoadNumPixels,
pxPerWorkItemX, pxPerWorkItemY, privDataWidth, borderMap[borderType],
isolated ? "BORDER_ISOLATED" : "NO_BORDER_ISOLATED",
privDataWidth / pxLoadNumPixels, pxPerWorkItemY + ksize.height - 1,
ocl::typeToStr(type), ocl::typeToStr(sdepth), ocl::typeToStr(dtype),
ocl::typeToStr(ddepth), ocl::typeToStr(wtype), ocl::typeToStr(wdepth),
ocl::convertTypeStr(sdepth, wdepth, cn, cvt[0]),
ocl::convertTypeStr(wdepth, ddepth, cn, cvt[1]), kerStr.c_str());
if (!k.create("filter2DSmall", cv::ocl::imgproc::filter2DSmall_oclsrc, build_options))
return false;
}
else
{
localsize = localsize_general;
std::vector<float> kernelMatDataFloat;
int kernel_size_y2_aligned = _prepareKernelFilter2D<float>(kernelMatDataFloat, kernelMat);
String kerStr = ocl::kernelToStr(kernelMatDataFloat, CV_32F);
for ( ; ; )
{
size_t BLOCK_SIZE = tryWorkItems;
while (BLOCK_SIZE > 32 && BLOCK_SIZE >= (size_t)ksize.width * 2 && BLOCK_SIZE > (size_t)sz.width * 2)
BLOCK_SIZE /= 2;
if ((size_t)ksize.width > BLOCK_SIZE)
return false;
int requiredTop = anchor.y;
int requiredLeft = (int)BLOCK_SIZE; // not this: anchor.x;
int requiredBottom = ksize.height - 1 - anchor.y;
int requiredRight = (int)BLOCK_SIZE; // not this: ksize.width - 1 - anchor.x;
int h = isolated ? sz.height : wholeSize.height;
int w = isolated ? sz.width : wholeSize.width;
bool extra_extrapolation = h < requiredTop || h < requiredBottom || w < requiredLeft || w < requiredRight;
if ((w < ksize.width) || (h < ksize.height))
return false;
String opts = format("-D LOCAL_SIZE=%d -D cn=%d "
"-D ANCHOR_X=%d -D ANCHOR_Y=%d -D KERNEL_SIZE_X=%d -D KERNEL_SIZE_Y=%d "
"-D KERNEL_SIZE_Y2_ALIGNED=%d -D %s -D %s -D %s%s%s "
"-D srcT=%s -D srcT1=%s -D dstT=%s -D dstT1=%s -D WT=%s -D WT1=%s "
"-D convertToWT=%s -D convertToDstT=%s",
(int)BLOCK_SIZE, cn, anchor.x, anchor.y,
ksize.width, ksize.height, kernel_size_y2_aligned, borderMap[borderType],
extra_extrapolation ? "EXTRA_EXTRAPOLATION" : "NO_EXTRA_EXTRAPOLATION",
isolated ? "BORDER_ISOLATED" : "NO_BORDER_ISOLATED",
doubleSupport ? " -D DOUBLE_SUPPORT" : "", kerStr.c_str(),
ocl::typeToStr(type), ocl::typeToStr(sdepth), ocl::typeToStr(dtype),
ocl::typeToStr(ddepth), ocl::typeToStr(wtype), ocl::typeToStr(wdepth),
ocl::convertTypeStr(sdepth, wdepth, cn, cvt[0]),
ocl::convertTypeStr(wdepth, ddepth, cn, cvt[1]));
localsize[0] = BLOCK_SIZE;
globalsize[0] = DIVUP(sz.width, BLOCK_SIZE - (ksize.width - 1)) * BLOCK_SIZE;
globalsize[1] = sz.height;
if (!k.create("filter2D", cv::ocl::imgproc::filter2D_oclsrc, opts))
return false;
size_t kernelWorkGroupSize = k.workGroupSize();
if (localsize[0] <= kernelWorkGroupSize)
break;
if (BLOCK_SIZE < kernelWorkGroupSize)
return false;
tryWorkItems = kernelWorkGroupSize;
}
}
_dst.create(sz, dtype);
UMat dst = _dst.getUMat();
int srcOffsetX = (int)((src.offset % src.step) / src.elemSize());
int srcOffsetY = (int)(src.offset / src.step);
int srcEndX = (isolated ? (srcOffsetX + sz.width) : wholeSize.width);
int srcEndY = (isolated ? (srcOffsetY + sz.height) : wholeSize.height);
k.args(ocl::KernelArg::PtrReadOnly(src), (int)src.step, srcOffsetX, srcOffsetY,
srcEndX, srcEndY, ocl::KernelArg::WriteOnly(dst), (float)delta);
return k.run(2, globalsize, localsize, false);
}