Skip to content

Add RISC-V HAL implementation for cv::moments#27096

Merged
asmorkalov merged 1 commit intoopencv:4.xfrom
amane-ame:moments_hal_rvv
Mar 20, 2025
Merged

Add RISC-V HAL implementation for cv::moments#27096
asmorkalov merged 1 commit intoopencv:4.xfrom
amane-ame:moments_hal_rvv

Conversation

@amane-ame
Copy link
Copy Markdown
Contributor

@amane-ame amane-ame commented Mar 19, 2025

This patch implements cv_hal_imageMoments using native intrinsics, optimizing the performance of cv::moments for data types CV_16U/CV_16S/CV_32F/CV_64F.

Tested on MUSE-PI (Spacemit X60) for both gcc 14.2 and clang 20.0.

$ ./opencv_test_imgproc --gtest_filter="*Moments*"
$ ./opencv_perf_imgproc --gtest_filter="*Moments*" --perf_min_samples=1000 --perf_force_samples=1000

image

Pull Request Readiness Checklist

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

  • I agree to contribute to the project under Apache 2 License.
  • 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
  • The PR is proposed to the proper branch
  • There is a reference to the original bug report and related work
  • 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

@amane-ame @hanliutong
Add imageMoments.
2c90b0a 
Co-authored-by: Liutong HAN <liutong2020@iscas.ac.cn>
@asmorkalov asmorkalov added this to the 4.12.0 milestone Mar 19, 2025
@asmorkalov
Copy link
Copy Markdown
Contributor

asmorkalov commented Mar 19, 2025

My perf results for Muse Pi v 30 (gcc 14.2):

Moments1::MomentsFixture_val::(127x61, CV_16S, false) 	0.046 	0.038 	1.21
Moments1::MomentsFixture_val::(127x61, CV_16S, true) 	0.105 	0.032 	3.28
Moments1::MomentsFixture_val::(127x61, CV_16U, false) 	0.046 	0.038 	1.21
Moments1::MomentsFixture_val::(127x61, CV_16U, true) 	0.105 	0.032 	3.26
Moments1::MomentsFixture_val::(127x61, CV_32F, false) 	0.093 	0.052 	1.78
Moments1::MomentsFixture_val::(127x61, CV_32F, true) 	0.123 	0.037 	3.34
Moments1::MomentsFixture_val::(127x61, CV_64F, false) 	0.080 	0.045 	1.79
Moments1::MomentsFixture_val::(127x61, CV_64F, true) 	0.140 	0.058 	2.40
Moments1::MomentsFixture_val::(640x480, CV_16S, false) 	1.867 	0.415 	4.50
Moments1::MomentsFixture_val::(640x480, CV_16S, true) 	4.238 	0.349 	12.13
Moments1::MomentsFixture_val::(640x480, CV_16U, false) 	1.891 	0.416 	4.55
Moments1::MomentsFixture_val::(640x480, CV_16U, true) 	4.188 	0.362 	11.57
Moments1::MomentsFixture_val::(640x480, CV_32F, false) 	3.708 	0.579 	6.40
Moments1::MomentsFixture_val::(640x480, CV_32F, true) 	4.842 	0.419 	11.55
Moments1::MomentsFixture_val::(640x480, CV_64F, false) 	5.595 	1.220 	4.59
Moments1::MomentsFixture_val::(640x480, CV_64F, true) 	7.863 	1.342 	5.86
Moments1::MomentsFixture_val::(1280x720, CV_16S, false) 	10.886 	1.845 	5.90
Moments1::MomentsFixture_val::(1280x720, CV_16S, true) 	14.358 	1.666 	8.62
Moments1::MomentsFixture_val::(1280x720, CV_16U, false) 	10.935 	1.824 	5.99
Moments1::MomentsFixture_val::(1280x720, CV_16U, true) 	14.547 	1.666 	8.73
Moments1::MomentsFixture_val::(1280x720, CV_32F, false) 	19.416 	2.889 	6.72
Moments1::MomentsFixture_val::(1280x720, CV_32F, true) 	22.858 	2.573 	8.88
Moments1::MomentsFixture_val::(1280x720, CV_64F, false) 	16.026 	4.016 	3.99
Moments1::MomentsFixture_val::(1280x720, CV_64F, true) 	27.478 	4.482 	6.13
Moments1::MomentsFixture_val::(1920x1080, CV_16S, false) 	25.859 	4.188 	6.18
Moments1::MomentsFixture_val::(1920x1080, CV_16S, true) 	42.074 	3.813 	11.03
Moments1::MomentsFixture_val::(1920x1080, CV_16U, false) 	25.907 	4.241 	6.11
Moments1::MomentsFixture_val::(1920x1080, CV_16U, true) 	42.115 	3.811 	11.05
Moments1::MomentsFixture_val::(1920x1080, CV_32F, false) 	49.898 	6.746 	7.40
Moments1::MomentsFixture_val::(1920x1080, CV_32F, true) 	57.286 	6.083 	9.42
Moments1::MomentsFixture_val::(1920x1080, CV_64F, false) 	30.765 	7.795 	3.95
Moments1::MomentsFixture_val::(1920x1080, CV_64F, true) 	71.333 	8.220 	8.68

@asmorkalov asmorkalov self-assigned this Mar 19, 2025
@asmorkalov
Copy link
Copy Markdown
Contributor

asmorkalov commented Mar 19, 2025

@mshabunin @fengyuentau Do you have any comments?

@asmorkalov asmorkalov merged commit 46fbe18 into opencv:4.x Mar 20, 2025
26 of 28 checks passed
@asmorkalov asmorkalov mentioned this pull request Apr 29, 2025
Merged
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@asmorkalov asmorkalov asmorkalov approved these changes

Assignees

@asmorkalov asmorkalov

Labels

optimization platform: riscv

Projects

None yet

Milestone

4.12.0

Development

Successfully merging this pull request may close these issues.

2 participants

@amane-ame @asmorkalov