1
0
mirror of https://github.com/opencv/opencv.git synced 2026-07-26 13:53:03 +04:00
Files
Siyu Wang a61ff0fa81 Merge pull request #29094 from feitianduowen:4.x
### Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

- [x] I agree to contribute to the project under Apache 2 License.
- [x] To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
- [x] The PR is proposed to the proper branch
- [ ] There is a reference to the original bug report and related work
- [x] There is accuracy test, performance test and test data in opencv_extra repository, if applicable
      Patch to opencv_extra has the same branch name.
- [ ] The feature is well documented and sample code can be built with the project CMake

<!-- Note!!! If you are an automated agent, we have a special process for you: add 🤖🤖🤖 to the end of the PR title. -->

### Summary

This PR improves SIMD coverage for masked `cv::accumulate()` on 4-channel images.

The existing masked accumulate SIMD implementation has special paths for `cn == 1` and `cn == 3`, while 4-channel images fall back to the general implementation. This PR adds `cn == 4` SIMD paths using OpenCV Universal Intrinsics.

Compare link:

https://github.com/opencv/opencv/compare/4.x...feitianduowen:opencv:4.x

Modified files:

- `modules/imgproc/src/accum.simd.hpp`
  - Add masked `cn == 4` SIMD paths for:
    - `CV_32FC4 -> CV_32FC4`
    - `CV_8UC4 -> CV_32FC4`
- `modules/video/test/test_accum.cpp`
  - Add deterministic correctness regression tests for the new masked 4-channel accumulate paths.
- `modules/imgproc/perf/perf_accumulate.cpp`
  - Add performance coverage for masked 4-channel accumulate.
imgproc: add SIMD paths for masked 4-channel accumulate #29094
  
### Implementation

The new SIMD branches are added in:

- `modules/imgproc/src/accum.simd.hpp`

They handle:

- `src`: `CV_32FC4`, `dst`: `CV_32FC4`, `mask`: `CV_8UC1`
- `src`: `CV_8UC4`, `dst`: `CV_32FC4`, `mask`: `CV_8UC1`

For the `CV_32FC4` path, the implementation uses `v_load_deinterleave`, applies the pixel mask to all 4 channels, accumulates into `dst`, and stores the result with `v_store_interleave`.

For the `CV_8UC4 -> CV_32FC4` path, the implementation loads 4 interleaved uchar channels, applies the pixel mask, widens the values to float, accumulates into `dst`, and stores the results back as interleaved `CV_32FC4`.

The masked SIMD block is guarded with:

```cpp
if (x <= len - cVectorWidth)
{
    ...
}
```
This check ensures that SIMD setup and vector operations are only executed when at least one full vector iteration can run. For very small inputs or tail-only cases, the function skips SIMD initialization and lets the existing scalar fallback handle the data through `acc_general_(src, dst, mask, len, cn, x)`.

This keeps tail handling unchanged and avoids unnecessary SIMD initialization when the vector loop would be skipped.

### Accuracy tests

Added a deterministic regression test in:

* `modules/video/test/test_accum.cpp`

New test:

```bash
Video_Acc.accuracy_32FC4_masked_cn4
Video_Acc.accuracy_8UC4_to_32FC4_masked_cn4
```

Test commands:

```bash
opencv_test_video --gtest_filter=Video_Acc.accuracy_32FC4_masked_cn4
opencv_test_video --gtest_filter=Video_Acc.accuracy_8UC4_to_32FC4_masked_cn4
opencv_test_video --gtest_filter="Video_Acc.*:Video_AccSquared.*:Video_AccProduct.*:Video_RunningAvg.*"
```
The filtered accumulate-related tests passed locally.

I added dedicated deterministic tests instead of only extending the existing randomized accumulate tests because this PR only changes the masked `cv::accumulate()` paths for specific 4-channel type combinations. It does not add `cn == 4` SIMD coverage for `accumulateSquare`, `accumulateProduct`, or `accumulateWeighted`.

The existing base accumulate tests are shared by several accumulation functions. Extending the common randomized channel selection to include `cn == 4` would also affect tests for functions that are not optimized by this PR. The new deterministic tests directly cover the modified paths:

* `cv::accumulate(src, dst, mask)`
* `CV_32FC4 -> CV_32FC4`
* `CV_8UC4 -> CV_32FC4`
* `mask`: `CV_8UC1`

The test also covers small sizes and non-vector-multiple sizes, so both the new SIMD path and the existing scalar tail fallback are exercised.

### Performance tests

Added performance coverage in:

* `modules/imgproc/perf/perf_accumulate.cpp`

Perf command:

```bash
opencv_perf_imgproc --gtest_filter="*AccumulateMask32FC4*" --perf_min_samples=1000 --perf_force_samples=1000
opencv_perf_imgproc --gtest_filter="*AccumulateMask8UC4To32FC4*" --perf_min_samples=1000 --perf_force_samples=1000
```

Test environment:

* Release build
* AVX2 enabled
* IPP disabled
* OpenCL disabled
* Windows MinGW build

Performance results:

`CV_32FC4 -> CV_32FC4`

| Size      | Before median | After median | Speedup |
| --------- | ------------- | ------------ | ------- |
| 1920x1080 | 3.07 ms       | 2.74 ms      | 1.12x   |
| 1280x720  | 0.84 ms       | 0.73 ms      | 1.15x   |
| 640x480   | 0.17 ms       | 0.16 ms      | 1.06x   |
| 320x240   | 0.04 ms       | 0.04 ms      | ~1.00x  |

`CV_8UC4 -> CV_32FC4`

| Size      | Before median | After median | Speedup |
| --------- | ------------- | ------------ | ------- |
| 1920x1080 | 6.09 ms       | 1.75 ms      | 3.48x   |
| 1280x720  | 2.69 ms       | 0.82 ms      | 3.28x   |
| 640x480   | 0.96 ms       | 0.28 ms      | 3.43x   |
| 320x240   | 0.24 ms       | 0.06 ms      | 4.00x   |
| 127x61    | 0.02 ms       | 0.01 ms      | 2.00x   |

### Build and test commands

The following commands were run from Windows `cmd.exe`.

```sh
cmake -S . -B build_release -G Ninja -DCMAKE_BUILD_TYPE=Release -DBUILD_TESTS=ON -DBUILD_PERF_TESTS=OFF -DBUILD_EXAMPLES=OFF -DWITH_IPP=OFF -DWITH_OPENCL=OFF -DCMAKE_C_FLAGS=-mstackrealign -DCMAKE_CXX_FLAGS=-mstackrealign

cmake --build build_release --target opencv_test_video -j 4

build_release\bin\opencv_test_video.exe --gtest_filter=Video_Acc.accuracy_32FC4_masked_cn4
build_release\bin\opencv_test_video.exe --gtest_filter=Video_Acc.accuracy_8UC4_to_32FC4_masked_cn4
build_release\bin\opencv_test_video.exe --gtest_filter=Video_Acc.*:Video_AccSquared.*:Video_AccProduct.*:Video_RunningAvg.*
```

Accuracy test commands passed locally.

For the performance comparison, I kept the new perf test in `modules/imgproc/perf/perf_accumulate.cpp` and only reverted `modules/imgproc/src/accum.simd.hpp` to measure the baseline. Then I restored the SIMD patch and measured the optimized version.

Baseline measurement:

```sh
git diff -- modules/imgproc/src/accum.simd.hpp > accum_cn4_simd.patch
git checkout -- modules/imgproc/src/accum.simd.hpp

mkdir perf_logs

cmake -S . -B build_perf_before -G Ninja -DCMAKE_BUILD_TYPE=Release -DBUILD_TESTS=OFF -DBUILD_PERF_TESTS=ON -DBUILD_EXAMPLES=OFF -DWITH_IPP=OFF -DWITH_OPENCL=OFF -DCMAKE_C_FLAGS=-mstackrealign -DCMAKE_CXX_FLAGS=-mstackrealign > perf_logs\before_cmake_config.txt 2>&1
cmake --build build_perf_before --target opencv_perf_imgproc -j 4 > perf_logs\before_build.txt 2>&1

build_perf_before\bin\opencv_perf_imgproc.exe --gtest_filter=*AccumulateMask32FC4* --perf_min_samples=1000 --perf_force_samples=1000 > perf_logs\before_perf.txt 2>&1
```

```sh
git diff -- modules/imgproc/src/accum.simd.hpp > accum_cn4_u8_simd.patch
git checkout -- modules/imgproc/src/accum.simd.hpp

cmake -S . -B build_perf_before_u8 -G Ninja -DCMAKE_BUILD_TYPE=Release -DBUILD_TESTS=OFF -DBUILD_PERF_TESTS=ON -DBUILD_EXAMPLES=OFF -DWITH_IPP=OFF -DWITH_OPENCL=OFF -DCMAKE_C_FLAGS=-mstackrealign -DCMAKE_CXX_FLAGS=-mstackrealign
cmake --build build_perf_before_u8 --target opencv_perf_imgproc -j 4
build_perf_before_u8\bin\opencv_perf_imgproc.exe --gtest_filter=*AccumulateMask8UC4To32FC4* --perf_min_samples=1000 --perf_force_samples=1000 > perf_logs\before_perf1.txt 2>&1
```

Optimized measurement:

```sh
git apply accum_cn4_simd.patch

cmake -S . -B build_perf_after -G Ninja -DCMAKE_BUILD_TYPE=Release -DBUILD_TESTS=OFF -DBUILD_PERF_TESTS=ON -DBUILD_EXAMPLES=OFF -DWITH_IPP=OFF -DWITH_OPENCL=OFF -DCMAKE_C_FLAGS=-mstackrealign -DCMAKE_CXX_FLAGS=-mstackrealign > perf_logs\after_cmake_config.txt 2>&1
cmake --build build_perf_after --target opencv_perf_imgproc -j 4 > perf_logs\after_build.txt 2>&1
build_perf_after\bin\opencv_perf_imgproc.exe --gtest_filter=*AccumulateMask32FC4* --perf_min_samples=500 --perf_force_samples=500 > perf_logs\after_perf.txt 2>&1
```

```sh
git apply accum_cn4_u8_simd.patch

cmake -S . -B build_perf_after_u8 -G Ninja -DCMAKE_BUILD_TYPE=Release -DBUILD_TESTS=OFF -DBUILD_PERF_TESTS=ON -DBUILD_EXAMPLES=OFF -DWITH_IPP=OFF -DWITH_OPENCL=OFF -DCMAKE_C_FLAGS=-mstackrealign -DCMAKE_CXX_FLAGS=-mstackrealign
cmake --build build_perf_after_u8 --target opencv_perf_imgproc -j 4
build_perf_after_u8\bin\opencv_perf_imgproc.exe --gtest_filter=*AccumulateMask8UC4To32FC4* --perf_min_samples=1000 --perf_force_samples=1000 > perf_logs\after_perf1.txt 2>&1
```

Performance comparison was measured by keeping the new perf tests and only reverting modules/imgproc/src/accum.simd.hpp for the baseline run.

`-mstackrealign` was used for the MinGW Windows build to avoid stack-alignment issues with AVX code generation during local testing.

### Notes

The implementation keeps the existing scalar fallback path unchanged. SIMD is used only for the newly covered masked `cn == 4` vectorizable part, and any remaining tail elements are still handled by `acc_general_()`.

The improvement is most visible on larger images. Small images are dominated by overhead and do not always show meaningful speedup.
2026-05-28 12:44:16 +03:00

361 lines
10 KiB
C++

/*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.
//
//
// Intel License Agreement
// For Open Source Computer Vision Library
//
// Copyright (C) 2000, Intel Corporation, 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 Intel Corporation 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 "test_precomp.hpp"
#include "opencv2/imgproc/imgproc_c.h"
namespace opencv_test { namespace {
class CV_AccumBaseTest : public cvtest::ArrayTest
{
public:
CV_AccumBaseTest();
protected:
void get_test_array_types_and_sizes( int test_case_idx, vector<vector<Size> >& sizes, vector<vector<int> >& types );
double get_success_error_level( int test_case_idx, int i, int j );
double alpha;
};
CV_AccumBaseTest::CV_AccumBaseTest()
{
test_array[INPUT].push_back(NULL);
test_array[INPUT_OUTPUT].push_back(NULL);
test_array[REF_INPUT_OUTPUT].push_back(NULL);
test_array[MASK].push_back(NULL);
optional_mask = true;
element_wise_relative_error = false;
} // ctor
void CV_AccumBaseTest::get_test_array_types_and_sizes( int test_case_idx,
vector<vector<Size> >& sizes, vector<vector<int> >& types )
{
RNG& rng = ts->get_rng();
int depth = cvtest::randInt(rng) % 4, cn = cvtest::randInt(rng) & 1 ? 3 : 1;
int accdepth = (int)(cvtest::randInt(rng) % 2 + 1);
int i, input_count = (int)test_array[INPUT].size();
cvtest::ArrayTest::get_test_array_types_and_sizes( test_case_idx, sizes, types );
depth = depth == 0 ? CV_8U : depth == 1 ? CV_16U : depth == 2 ? CV_32F : CV_64F;
accdepth = accdepth == 1 ? CV_32F : CV_64F;
accdepth = MAX(accdepth, depth);
for( i = 0; i < input_count; i++ )
types[INPUT][i] = CV_MAKETYPE(depth,cn);
types[INPUT_OUTPUT][0] = types[REF_INPUT_OUTPUT][0] = CV_MAKETYPE(accdepth,cn);
alpha = cvtest::randReal(rng);
}
double CV_AccumBaseTest::get_success_error_level( int /*test_case_idx*/, int /*i*/, int /*j*/ )
{
return test_mat[INPUT_OUTPUT][0].depth() < CV_64F ||
test_mat[INPUT][0].depth() == CV_32F ? FLT_EPSILON*100 : DBL_EPSILON*1000;
}
/// acc
class CV_AccTest : public CV_AccumBaseTest
{
public:
CV_AccTest() { }
protected:
void run_func();
void prepare_to_validation( int );
};
void CV_AccTest::run_func(void)
{
cvAcc( test_array[INPUT][0], test_array[INPUT_OUTPUT][0], test_array[MASK][0] );
}
void CV_AccTest::prepare_to_validation( int )
{
const Mat& src = test_mat[INPUT][0];
Mat& dst = test_mat[REF_INPUT_OUTPUT][0];
const Mat& mask = test_array[MASK][0] ? test_mat[MASK][0] : Mat();
Mat temp;
cvtest::add( src, 1, dst, 1, cvScalarAll(0.), temp, dst.type() );
cvtest::copy( temp, dst, mask );
}
/// square acc
class CV_SquareAccTest : public CV_AccumBaseTest
{
public:
CV_SquareAccTest();
protected:
void run_func();
void prepare_to_validation( int );
};
CV_SquareAccTest::CV_SquareAccTest()
{
}
void CV_SquareAccTest::run_func()
{
cvSquareAcc( test_array[INPUT][0], test_array[INPUT_OUTPUT][0], test_array[MASK][0] );
}
void CV_SquareAccTest::prepare_to_validation( int )
{
const Mat& src = test_mat[INPUT][0];
Mat& dst = test_mat[REF_INPUT_OUTPUT][0];
const Mat& mask = test_array[MASK][0] ? test_mat[MASK][0] : Mat();
Mat temp;
cvtest::convert( src, temp, dst.type() );
cvtest::multiply( temp, temp, temp, 1 );
cvtest::add( temp, 1, dst, 1, cvScalarAll(0.), temp, dst.depth() );
cvtest::copy( temp, dst, mask );
}
/// multiply acc
class CV_MultiplyAccTest : public CV_AccumBaseTest
{
public:
CV_MultiplyAccTest();
protected:
void run_func();
void prepare_to_validation( int );
};
CV_MultiplyAccTest::CV_MultiplyAccTest()
{
test_array[INPUT].push_back(NULL);
}
void CV_MultiplyAccTest::run_func()
{
cvMultiplyAcc( test_array[INPUT][0], test_array[INPUT][1],
test_array[INPUT_OUTPUT][0], test_array[MASK][0] );
}
void CV_MultiplyAccTest::prepare_to_validation( int )
{
const Mat& src1 = test_mat[INPUT][0];
const Mat& src2 = test_mat[INPUT][1];
Mat& dst = test_mat[REF_INPUT_OUTPUT][0];
const Mat& mask = test_array[MASK][0] ? test_mat[MASK][0] : Mat();
Mat temp1, temp2;
cvtest::convert( src1, temp1, dst.type() );
cvtest::convert( src2, temp2, dst.type() );
cvtest::multiply( temp1, temp2, temp1, 1 );
cvtest::add( temp1, 1, dst, 1, cvScalarAll(0.), temp1, dst.depth() );
cvtest::copy( temp1, dst, mask );
}
/// running average
class CV_RunningAvgTest : public CV_AccumBaseTest
{
public:
CV_RunningAvgTest();
protected:
void run_func();
void prepare_to_validation( int );
};
CV_RunningAvgTest::CV_RunningAvgTest()
{
}
void CV_RunningAvgTest::run_func()
{
cvRunningAvg( test_array[INPUT][0], test_array[INPUT_OUTPUT][0],
alpha, test_array[MASK][0] );
}
void CV_RunningAvgTest::prepare_to_validation( int )
{
const Mat& src = test_mat[INPUT][0];
Mat& dst = test_mat[REF_INPUT_OUTPUT][0];
Mat temp;
const Mat& mask = test_array[MASK][0] ? test_mat[MASK][0] : Mat();
double a[1], b[1];
int accdepth = test_mat[INPUT_OUTPUT][0].depth();
CvMat A = cvMat(1,1,accdepth,a), B = cvMat(1,1,accdepth,b);
cvSetReal1D( &A, 0, alpha);
cvSetReal1D( &B, 0, 1 - cvGetReal1D(&A, 0));
cvtest::convert( src, temp, dst.type() );
cvtest::add( src, cvGetReal1D(&A, 0), dst, cvGetReal1D(&B, 0), cvScalarAll(0.), temp, temp.depth() );
cvtest::copy( temp, dst, mask );
}
TEST(Video_Acc, accuracy) { CV_AccTest test; test.safe_run(); }
TEST(Video_AccSquared, accuracy) { CV_SquareAccTest test; test.safe_run(); }
TEST(Video_AccProduct, accuracy) { CV_MultiplyAccTest test; test.safe_run(); }
TEST(Video_RunningAvg, accuracy) { CV_RunningAvgTest test; test.safe_run(); }
typedef testing::TestWithParam<tuple<Size, int, int> > Video_Acc_Cn4;
TEST_P(Video_Acc_Cn4, accuracy)
{
const Size size = get<0>(GetParam());
const int pattern = get<1>(GetParam());
const int srcType = get<2>(GetParam());
RNG& rng = theRNG();
Mat src(size, srcType);
Mat dst(size, CV_32FC4);
Mat mask(size, CV_8UC1);
if (srcType == CV_8UC4)
rng.fill(src, RNG::UNIFORM, Scalar::all(0), Scalar::all(256));
else
rng.fill(src, RNG::UNIFORM, Scalar::all(-10.0), Scalar::all(10.0));
rng.fill(dst, RNG::UNIFORM, Scalar::all(-1000.0), Scalar::all(1000.0));
for (int y = 0; y < mask.rows; ++y)
{
uchar* row = mask.ptr<uchar>(y);
for (int x = 0; x < mask.cols; ++x)
{
switch (pattern)
{
case 0:
row[x] = 0;
break;
case 1:
row[x] = 255;
break;
case 2:
row[x] = ((x + y) % 2) ? 255 : 0;
break;
case 3:
row[x] = ((x * 13 + y * 7) % 5) ? 255 : 0;
break;
default:
row[x] = ((x * 17 + y * 11) % 3) ? 255 : 0;
break;
}
}
}
Mat dstRef = dst.clone();
if (srcType == CV_32FC4)
{
for (int y = 0; y < src.rows; ++y)
{
const Vec4f* srcRow = src.ptr<Vec4f>(y);
Vec4f* dstRefRow = dstRef.ptr<Vec4f>(y);
const uchar* maskRow = mask.ptr<uchar>(y);
for (int x = 0; x < src.cols; ++x)
{
if (maskRow[x])
{
for (int c = 0; c < 4; ++c)
dstRefRow[x][c] += srcRow[x][c];
}
}
}
}
else
{
CV_Assert(srcType == CV_8UC4);
for (int y = 0; y < src.rows; ++y)
{
const Vec4b* srcRow = src.ptr<Vec4b>(y);
Vec4f* dstRefRow = dstRef.ptr<Vec4f>(y);
const uchar* maskRow = mask.ptr<uchar>(y);
for (int x = 0; x < src.cols; ++x)
{
if (maskRow[x])
{
for (int c = 0; c < 4; ++c)
dstRefRow[x][c] += static_cast<float>(srcRow[x][c]);
}
}
}
}
cv::accumulate(src, dst, mask);
const double err = cv::norm(dst, dstRef, NORM_INF);
EXPECT_EQ(0.0, err)
<< "size=" << size
<< ", pattern=" << pattern
<< ", srcType=" << srcType;
}
INSTANTIATE_TEST_CASE_P(Accumulate,
Video_Acc_Cn4,
testing::Combine(
testing::Values(Size(1, 1),
Size(3, 5),
Size(17, 7),
Size(37, 19),
Size(128, 16),
Size(641, 37)),
testing::Values(0, 1, 2, 3, 4),
testing::Values(CV_32FC4, CV_8UC4)));
}} // namespace