Skip to content

O
opencv
Commit d13a6b74 authored by Vladislav Vinogradov's avatar Vladislav Vinogradov
Browse files
Options

Fixed bug with submatrix in device::transform

parent dab35867
Show whitespace changes
Inline Side-by-side
Showing with 64 additions and 23 deletions
modules/core/include/opencv2/core/gpumat.hpp
View file @ d13a6b74
......@@ -448,7 +448,7 @@ namespace cv { namespace gpu
{
int area = rows * cols;
if (!m.isContinuous() || m.type() != type || m.size().area() != area)
m.create(1, area, type);
ensureSizeIsEnough(1, area, type, m);
m = m.reshape(0, rows);
}
......
modules/gpu/src/cuda/element_operations.cu
View file @ d13a6b74
......@@ -1058,12 +1058,12 @@ namespace cv { namespace gpu { namespace device
::cv::gpu::device::transform((DevMem2D_<T>)src1, (DevMem2D_<T>)src2, (DevMem2D_<T>)dst, Absdiff<T>(), stream);
}
//template void absdiff_gpu<uchar >(const DevMem2Db& src1, const DevMem2Db& src2, const DevMem2Db& dst, cudaStream_t stream);
template void absdiff_gpu<uchar >(const DevMem2Db& src1, const DevMem2Db& src2, const DevMem2Db& dst, cudaStream_t stream);
template void absdiff_gpu<schar >(const DevMem2Db& src1, const DevMem2Db& src2, const DevMem2Db& dst, cudaStream_t stream);
template void absdiff_gpu<ushort>(const DevMem2Db& src1, const DevMem2Db& src2, const DevMem2Db& dst, cudaStream_t stream);
template void absdiff_gpu<short >(const DevMem2Db& src1, const DevMem2Db& src2, const DevMem2Db& dst, cudaStream_t stream);
//template void absdiff_gpu<int >(const DevMem2Db& src1, const DevMem2Db& src2, const DevMem2Db& dst, cudaStream_t stream);
//template void absdiff_gpu<float >(const DevMem2Db& src1, const DevMem2Db& src2, const DevMem2Db& dst, cudaStream_t stream);
template void absdiff_gpu<int >(const DevMem2Db& src1, const DevMem2Db& src2, const DevMem2Db& dst, cudaStream_t stream);
template void absdiff_gpu<float >(const DevMem2Db& src1, const DevMem2Db& src2, const DevMem2Db& dst, cudaStream_t stream);
template void absdiff_gpu<double>(const DevMem2Db& src1, const DevMem2Db& src2, const DevMem2Db& dst, cudaStream_t stream);
template <typename T> struct AbsdiffScalar : unary_function<T, T>
......
modules/gpu/src/element_operations.cpp
View file @ d13a6b74
......@@ -159,7 +159,13 @@ void cv::gpu::add(const GpuMat& src1, const GpuMat& src2, GpuMat& dst, const Gpu
cudaStream_t stream = StreamAccessor::getStream(s);
if (mask.empty() && dst.type() == src1.type() && (src1.depth() == CV_8U || src1.depth() == CV_32S || src1.depth() == CV_32F))
bool useNpp =
mask.empty() &&
dst.type() == src1.type() &&
(src1.depth() == CV_8U || src1.depth() == CV_32S || src1.depth() == CV_32F) &&
(isAligned(src1.data, 16) && isAligned(src2.data, 16) && isAligned(dst.data, 16));
if (useNpp)
{
nppArithmCaller(src1, src2, dst, nppiAdd_8u_C1RSfs, nppiAdd_8u_C4RSfs, nppiAdd_32s_C1R, nppiAdd_32f_C1R, stream);
return;
......@@ -271,7 +277,13 @@ void cv::gpu::subtract(const GpuMat& src1, const GpuMat& src2, GpuMat& dst, cons
cudaStream_t stream = StreamAccessor::getStream(s);
if (mask.empty() && dst.type() == src1.type() && (src1.depth() == CV_8U || src1.depth() == CV_32S || src1.depth() == CV_32F))
bool useNpp =
mask.empty() &&
dst.type() == src1.type() &&
(src1.depth() == CV_8U || src1.depth() == CV_32S || src1.depth() == CV_32F) &&
(isAligned(src1.data, 16) && isAligned(src2.data, 16) && isAligned(dst.data, 16));
if (useNpp)
{
nppArithmCaller(src2, src1, dst, nppiSub_8u_C1RSfs, nppiSub_8u_C4RSfs, nppiSub_32s_C1R, nppiSub_32f_C1R, stream);
return;
......@@ -403,8 +415,13 @@ void cv::gpu::multiply(const GpuMat& src1, const GpuMat& src2, GpuMat& dst, doub
dst.create(src1.size(), CV_MAKE_TYPE(CV_MAT_DEPTH(dtype), src1.channels()));
bool useNpp =
scale == 1 &&
dst.type() == src1.type() &&
(src1.depth() == CV_8U || src1.depth() == CV_32S || src1.depth() == CV_32F) &&
(isAligned(src1.data, 16) && isAligned(src2.data, 16) && isAligned(dst.data, 16));
if (scale == 1 && dst.type() == src1.type() && (src1.depth() == CV_8U || src1.depth() == CV_32S || src1.depth() == CV_32F))
if (useNpp)
{
nppArithmCaller(src2, src1, dst, nppiMul_8u_C1RSfs, nppiMul_8u_C4RSfs, nppiMul_32s_C1R, nppiMul_32f_C1R, stream);
return;
......@@ -528,8 +545,13 @@ void cv::gpu::divide(const GpuMat& src1, const GpuMat& src2, GpuMat& dst, double
dst.create(src1.size(), CV_MAKE_TYPE(CV_MAT_DEPTH(dtype), src1.channels()));
bool useNpp =
scale == 1 &&
dst.type() == src1.type() &&
(src1.depth() == CV_8U || src1.depth() == CV_32S || src1.depth() == CV_32F) &&
(isAligned(src1.data, 16) && isAligned(src2.data, 16) && isAligned(dst.data, 16));
if (scale == 1 && dst.type() == src1.type() && (src1.depth() == CV_8U || src1.depth() == CV_32S || src1.depth() == CV_32F))
if (useNpp)
{
nppArithmCaller(src2, src1, dst, nppiDiv_8u_C1RSfs, nppiDiv_8u_C4RSfs, nppiDiv_32s_C1R, nppiDiv_32f_C1R, stream);
return;
......@@ -643,7 +665,7 @@ void cv::gpu::absdiff(const GpuMat& src1, const GpuMat& src2, GpuMat& dst, Strea
static const func_t funcs[] =
{
0/*absdiff_gpu<unsigned char>*/, absdiff_gpu<signed char>, absdiff_gpu<unsigned short>, absdiff_gpu<short>, 0/*absdiff_gpu<int>*/, 0/*absdiff_gpu<float>*/, absdiff_gpu<double>
absdiff_gpu<unsigned char>, absdiff_gpu<signed char>, absdiff_gpu<unsigned short>, absdiff_gpu<short>, absdiff_gpu<int>, absdiff_gpu<float>, absdiff_gpu<double>
};
CV_Assert(src1.size() == src2.size() && src1.type() == src2.type());
......@@ -656,7 +678,9 @@ void cv::gpu::absdiff(const GpuMat& src1, const GpuMat& src2, GpuMat& dst, Strea
sz.width = src1.cols * src1.channels();
sz.height = src1.rows;
if (src1.depth() == CV_8U && (src1.cols * src1.channels()) % 4 == 0)
bool aligned = isAligned(src1.data, 16) && isAligned(src2.data, 16) && isAligned(dst.data, 16);
if (aligned && src1.depth() == CV_8U && (src1.cols * src1.channels()) % 4 == 0)
{
NppStreamHandler h(stream);
......@@ -668,7 +692,7 @@ void cv::gpu::absdiff(const GpuMat& src1, const GpuMat& src2, GpuMat& dst, Strea
if (stream == 0)
cudaSafeCall( cudaDeviceSynchronize() );
}
else if (src1.depth() == CV_8U)
else if (aligned && src1.depth() == CV_8U)
{
NppStreamHandler h(stream);
......@@ -678,7 +702,7 @@ void cv::gpu::absdiff(const GpuMat& src1, const GpuMat& src2, GpuMat& dst, Strea
if (stream == 0)
cudaSafeCall( cudaDeviceSynchronize() );
}
else if (src1.depth() == CV_32S)
else if (aligned && src1.depth() == CV_32S)
{
NppStreamHandler h(stream);
......@@ -688,7 +712,7 @@ void cv::gpu::absdiff(const GpuMat& src1, const GpuMat& src2, GpuMat& dst, Strea
if (stream == 0)
cudaSafeCall( cudaDeviceSynchronize() );
}
else if (src1.depth() == CV_32F)
else if (aligned && src1.depth() == CV_32F)
{
NppStreamHandler h(stream);
......
modules/gpu/src/opencv2/gpu/device/common.hpp
View file @ d13a6b74
......@@ -67,6 +67,11 @@
namespace cv { namespace gpu
{
void error(const char *error_string, const char *file, const int line, const char *func);
template <typename T> static inline bool isAligned(const T* ptr, size_t size)
{
return reinterpret_cast<size_t>(ptr) % size == 0;
}
}}
static inline void ___cudaSafeCall(cudaError_t err, const char *file, const int line, const char *func = "")
......
modules/gpu/src/opencv2/gpu/device/detail/transform_detail.hpp
View file @ d13a6b74
......@@ -309,7 +309,7 @@ namespace cv { namespace gpu { namespace device
template<> struct TransformDispatcher<false>
{
template <typename T, typename D, typename UnOp, typename Mask>
static void call(const DevMem2D_<T>& src, const DevMem2D_<D>& dst, const UnOp& op, const Mask& mask, cudaStream_t stream)
static void call(DevMem2D_<T> src, DevMem2D_<D> dst, UnOp op, Mask mask, cudaStream_t stream)
{
typedef TransformFunctorTraits<UnOp> ft;
......@@ -324,7 +324,7 @@ namespace cv { namespace gpu { namespace device
}
template <typename T1, typename T2, typename D, typename BinOp, typename Mask>
static void call(const DevMem2D_<T1>& src1, const DevMem2D_<T2>& src2, const DevMem2D_<D>& dst, const BinOp& op, const Mask& mask, cudaStream_t stream)
static void call(DevMem2D_<T1> src1, DevMem2D_<T2> src2, DevMem2D_<D> dst, BinOp op, Mask mask, cudaStream_t stream)
{
typedef TransformFunctorTraits<BinOp> ft;
......@@ -341,12 +341,18 @@ namespace cv { namespace gpu { namespace device
template<> struct TransformDispatcher<true>
{
template <typename T, typename D, typename UnOp, typename Mask>
static void call(const DevMem2D_<T>& src, const DevMem2D_<D>& dst, const UnOp& op, const Mask& mask, cudaStream_t stream)
static void call(DevMem2D_<T> src, DevMem2D_<D> dst, UnOp op, Mask mask, cudaStream_t stream)
{
typedef TransformFunctorTraits<UnOp> ft;
StaticAssert<ft::smart_shift != 1>::check();
if (!isAligned(src.data, ft::smart_shift * sizeof(T)) || !isAligned(dst.data, ft::smart_shift * sizeof(D)))
{
TransformDispatcher<false>::call(src, dst, op, mask, stream);
return;
}
const dim3 threads(ft::smart_block_dim_x, ft::smart_block_dim_y, 1);
const dim3 grid(divUp(src.cols, threads.x * ft::smart_shift), divUp(src.rows, threads.y), 1);
......@@ -358,12 +364,18 @@ namespace cv { namespace gpu { namespace device
}
template <typename T1, typename T2, typename D, typename BinOp, typename Mask>
static void call(const DevMem2D_<T1>& src1, const DevMem2D_<T2>& src2, const DevMem2D_<D>& dst, const BinOp& op, const Mask& mask, cudaStream_t stream)
static void call(DevMem2D_<T1> src1, DevMem2D_<T2> src2, DevMem2D_<D> dst, BinOp op, Mask mask, cudaStream_t stream)
{
typedef TransformFunctorTraits<BinOp> ft;
StaticAssert<ft::smart_shift != 1>::check();
if (!isAligned(src1.data, ft::smart_shift * sizeof(T1)) || !isAligned(src2.data, ft::smart_shift * sizeof(T2)) || !isAligned(dst.data, ft::smart_shift * sizeof(D)))
{
TransformDispatcher<false>::call(src1, src2, dst, op, mask, stream);
return;
}
const dim3 threads(ft::smart_block_dim_x, ft::smart_block_dim_y, 1);
const dim3 grid(divUp(src1.cols, threads.x * ft::smart_shift), divUp(src1.rows, threads.y), 1);
......@@ -376,14 +388,14 @@ namespace cv { namespace gpu { namespace device
};
template <typename T, typename D, typename UnOp, typename Mask>
static void transform_caller(const DevMem2D_<T>& src, const DevMem2D_<D>& dst, const UnOp& op, const Mask& mask, cudaStream_t stream)
static inline void transform_caller(DevMem2D_<T> src, DevMem2D_<D> dst, UnOp op, Mask mask, cudaStream_t stream)
{
typedef TransformFunctorTraits<UnOp> ft;
TransformDispatcher<VecTraits<T>::cn == 1 && VecTraits<D>::cn == 1 && ft::smart_shift != 1>::call(src, dst, op, mask, stream);
}
template <typename T1, typename T2, typename D, typename BinOp, typename Mask>
static void transform_caller(const DevMem2D_<T1>& src1, const DevMem2D_<T2>& src2, const DevMem2D_<D>& dst, const BinOp& op, const Mask& mask, cudaStream_t stream)
static inline void transform_caller(DevMem2D_<T1> src1, DevMem2D_<T2> src2, DevMem2D_<D> dst, BinOp op, Mask mask, cudaStream_t stream)
{
typedef TransformFunctorTraits<BinOp> ft;
TransformDispatcher<VecTraits<T1>::cn == 1 && VecTraits<T2>::cn == 1 && VecTraits<D>::cn == 1 && ft::smart_shift != 1>::call(src1, src2, dst, op, mask, stream);
......
modules/gpu/src/opencv2/gpu/device/transform.hpp
View file @ d13a6b74
......@@ -50,25 +50,25 @@
namespace cv { namespace gpu { namespace device
{
template <typename T, typename D, typename UnOp>
void transform(const DevMem2D_<T>& src, const DevMem2D_<D>& dst, const UnOp& op, cudaStream_t stream = 0)
static inline void transform(DevMem2D_<T> src, DevMem2D_<D> dst, UnOp op, cudaStream_t stream = 0)
{
transform_detail::transform_caller(src, dst, op, WithOutMask(), stream);
}
template <typename T, typename D, typename UnOp>
void transform(const DevMem2D_<T>& src, const DevMem2D_<D>& dst, const PtrStepb& mask, const UnOp& op, cudaStream_t stream = 0)
static inline void transform(DevMem2D_<T> src, DevMem2D_<D> dst, PtrStepb mask, UnOp op, cudaStream_t stream = 0)
{
transform_detail::transform_caller(src, dst, op, SingleMask(mask), stream);
}
template <typename T1, typename T2, typename D, typename BinOp>
void transform(const DevMem2D_<T1>& src1, const DevMem2D_<T2>& src2, const DevMem2D_<D>& dst, const BinOp& op, cudaStream_t stream = 0)
static inline void transform(DevMem2D_<T1> src1, DevMem2D_<T2> src2, DevMem2D_<D> dst, BinOp op, cudaStream_t stream = 0)
{
transform_detail::transform_caller(src1, src2, dst, op, WithOutMask(), stream);
}
template <typename T1, typename T2, typename D, typename BinOp>
void transform(const DevMem2D_<T1>& src1, const DevMem2D_<T2>& src2, const DevMem2D_<D>& dst, const PtrStepb& mask, const BinOp& op, cudaStream_t stream = 0)
static inline void transform(DevMem2D_<T1> src1, DevMem2D_<T2> src2, DevMem2D_<D> dst, PtrStepb mask, BinOp op, cudaStream_t stream = 0)
{
transform_detail::transform_caller(src1, src2, dst, op, SingleMask(mask), stream);
}
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment