FastCV-based HAL for OpenCV acceleration 2ndpost-1 by quic-xuezha · Pull Request #26617 · opencv/opencv

quic-xuezha · 2024-12-12T06:21:17Z

Detailed description:

Add parallel support for cv_hal_sobel
Add cv_hal_gaussianBlurBinomial and parallel support.
Add cv_hal_addWeighted8u and parallel support
Add cv_hal_warpPerspective and parallel support

Requires binary from opencv/opencv_3rdparty#90
Related patch to opencv_contrib: opencv/opencv_contrib#3844

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

remove INTER_LINEAR support because not bit-exact

asmorkalov · 2024-12-12T09:36:14Z

@quic-xuezha Thanks a lot for the contribution! Could you enable FastCV by default in the PR to check it with CI. We will revert the change just before merge.

asmorkalov · 2024-12-12T09:37:39Z

There is no FastCV links & hashes update in CMake. Old library will be used.

3rdparty/fastcv/src/fastcv_hal_imgproc.cpp

asmorkalov · 2024-12-17T07:21:27Z

Linux configuration build issue:

-- FASTCV: Downloading fastcv_linux_aarch64_2024_12_11.tgz from https://raw.githubusercontent.com/opencv/opencv_3rdparty/1eb84499ee9996616529fc48b0bab5631f64c2dc/fastcv/fastcv_linux_aarch64_2024_12_11.tgz
CMake Warning at cmake/OpenCVDownload.cmake:271 (message):
  FASTCV: Hash mismatch: a194679b934516dc2d377bb33e71ae76
Call Stack (most recent call first):
  3rdparty/fastcv/fastcv.cmake:29 (ocv_download)
  cmake/OpenCVFindLibsPerf.cmake:187 (download_fastcv)
  CMakeLists.txt:797 (include)


CMake Warning at 3rdparty/fastcv/fastcv.cmake:40 (message):
  FastCV: package download failed!
Call Stack (most recent call first):
  cmake/OpenCVFindLibsPerf.cmake:187 (download_fastcv)
  CMakeLists.txt:797 (include)

quic-xuezha · 2024-12-17T07:26:54Z

will check and update the hash

quic-xuezha · 2024-12-17T07:42:15Z

updated, please try again

asmorkalov · 2024-12-17T11:07:34Z

@quic-xuezha Thanks a lot for the contribution. I squashed the commits in opencv_3rdparty repo and created branch fastcv/4.x_20241211 there. Please update links/hashes again and revert the change that enables FastCV by default. I'll merge the PR after it.

This reverts commit 521c49d.

quic-xuezha · 2024-12-18T03:14:58Z

Updated the commit id and reverted the change

FastCV-based HAL for OpenCV acceleration 2ndpost-2 #26619 ### Detailed description: - Add support for multiply 8u, 16s and 32f - Add support for cv_hal_pyrdown 8u - Add support for cv_hal_cvtBGRtoHSV and cv_hal_cvtBGRtoYUVApprox 8u Requires binary from [opencv/opencv_3rdparty#90](opencv/opencv_3rdparty#90) Depends on: [#26617](#26617) ### 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 - [ ] 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

FastCV-based HAL for OpenCV acceleration 2ndpost-3 #26621 ### Detailed description: - Add cv_hal_canny for Canny API Requires binary from [opencv/opencv_3rdparty#90](opencv/opencv_3rdparty#90) Depends on: [#26617](#26617) Depends on: [#26619](#26619) ### 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 - [ ] 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

FastCV Extension code for OpenCV 2ndpost-1 #3844 Depends on: [opencv/opencv#26617](opencv/opencv#26617) Requires binary from [opencv/opencv_3rdparty#90](opencv/opencv_3rdparty#90) ### 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 - [ ] 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

FastCV-based HAL for OpenCV acceleration 2ndpost-1 opencv#26617 ### Detailed description: - Add parallel support for cv_hal_sobel - Add cv_hal_gaussianBlurBinomial and parallel support. - Add cv_hal_addWeighted8u and parallel support - Add cv_hal_warpPerspective and parallel support Requires binary from [opencv/opencv_3rdparty#90](opencv/opencv_3rdparty#90) Related patch to opencv_contrib: [opencv/opencv_contrib#3844](opencv/opencv_contrib#3844) ### 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 - [ ] 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

FastCV-based HAL for OpenCV acceleration 2ndpost-2 opencv#26619 ### Detailed description: - Add support for multiply 8u, 16s and 32f - Add support for cv_hal_pyrdown 8u - Add support for cv_hal_cvtBGRtoHSV and cv_hal_cvtBGRtoYUVApprox 8u Requires binary from [opencv/opencv_3rdparty#90](opencv/opencv_3rdparty#90) Depends on: [opencv#26617](opencv#26617) ### 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 - [ ] 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

FastCV-based HAL for OpenCV acceleration 2ndpost-3 opencv#26621 ### Detailed description: - Add cv_hal_canny for Canny API Requires binary from [opencv/opencv_3rdparty#90](opencv/opencv_3rdparty#90) Depends on: [opencv#26617](opencv#26617) Depends on: [opencv#26619](opencv#26619) ### 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 - [ ] 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

FastCV-based HAL for OpenCV acceleration 2ndpost-1 opencv#26617 ### Detailed description: - Add parallel support for cv_hal_sobel - Add cv_hal_gaussianBlurBinomial and parallel support. - Add cv_hal_addWeighted8u and parallel support - Add cv_hal_warpPerspective and parallel support Requires binary from [opencv/opencv_3rdparty#90](opencv/opencv_3rdparty#90) Related patch to opencv_contrib: [opencv/opencv_contrib#3844](opencv/opencv_contrib#3844) ### 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 - [ ] 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

FastCV-based HAL for OpenCV acceleration 2ndpost-2 opencv#26619 ### Detailed description: - Add support for multiply 8u, 16s and 32f - Add support for cv_hal_pyrdown 8u - Add support for cv_hal_cvtBGRtoHSV and cv_hal_cvtBGRtoYUVApprox 8u Requires binary from [opencv/opencv_3rdparty#90](opencv/opencv_3rdparty#90) Depends on: [opencv#26617](opencv#26617) ### 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 - [ ] 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

FastCV-based HAL for OpenCV acceleration 2ndpost-3 opencv#26621 ### Detailed description: - Add cv_hal_canny for Canny API Requires binary from [opencv/opencv_3rdparty#90](opencv/opencv_3rdparty#90) Depends on: [opencv#26617](opencv#26617) Depends on: [opencv#26619](opencv#26619) ### 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 - [ ] 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

quic-xuezha added 6 commits December 12, 2024 11:03

Add cv_hal_gaussianBlurBinomial and parallel support

b533999

Add cv_hal_addWeighted8u and parallel support

55611c7

Add cv_hal_warpPerspective and parallel support

63500ed

perpare for 2nd post

0dd8d58

remove INTER_LINEAR support because not bit-exact

Add ROI and parallel support for sobel

712e780

change nStripes calculation strategy and fix some failed cases

c497057

quic-xuezha mentioned this pull request Dec 12, 2024

FastCV Extension code for OpenCV 2ndpost-1 opencv/opencv_contrib#3844

Merged

6 tasks

asmorkalov requested a review from savuor December 12, 2024 09:20

asmorkalov added optimization category: imgproc platform: arm ARM boards related issues: RPi, NVIDIA TK/TX, etc labels Dec 12, 2024

quic-xuezha and others added 2 commits December 12, 2024 15:16

update fastcv lib commit id and hash code

a987f33

Enable fastcv by default

521c49d

adsha-quic mentioned this pull request Dec 12, 2024

FastCV-based HAL for OpenCV acceleration 2ndpost-2 #26619

Merged

6 tasks

quic-apreetam mentioned this pull request Dec 12, 2024

FastCV-based HAL for OpenCV acceleration 2ndpost-3 #26621

Merged

6 tasks

update linux lib hash code

e4f74e5

asmorkalov reviewed Dec 13, 2024

View reviewed changes

3rdparty/fastcv/src/fastcv_hal_imgproc.cpp Outdated Show resolved Hide resolved

3rdparty/fastcv/src/fastcv_hal_imgproc.cpp Outdated Show resolved Hide resolved

asmorkalov self-assigned this Dec 16, 2024

remove memcpy and add epsilon for float comparison

80b35a3

update commit id for lib

a00a1db

asmorkalov approved these changes Dec 17, 2024

View reviewed changes

asmorkalov added this to the 4.11.0 milestone Dec 17, 2024

quic-xuezha and others added 2 commits December 18, 2024 08:01

update commit id

8763a7c

Revert "Enable fastcv by default"

17cc522

This reverts commit 521c49d.

asmorkalov merged commit 1c28a98 into opencv:4.x Dec 18, 2024

asmorkalov mentioned this pull request Jan 15, 2025

5.x merge 4.x #26768

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

FastCV-based HAL for OpenCV acceleration 2ndpost-1#26617

FastCV-based HAL for OpenCV acceleration 2ndpost-1#26617
asmorkalov merged 13 commits intoopencv:4.xfrom
CodeLinaro:xuezha_2ndPost

quic-xuezha commented Dec 12, 2024 •

edited

Loading

Uh oh!

asmorkalov commented Dec 12, 2024

Uh oh!

asmorkalov commented Dec 12, 2024

Uh oh!

Uh oh!

Uh oh!

asmorkalov commented Dec 17, 2024

Uh oh!

quic-xuezha commented Dec 17, 2024

Uh oh!

quic-xuezha commented Dec 17, 2024

Uh oh!

asmorkalov commented Dec 17, 2024

Uh oh!

quic-xuezha commented Dec 18, 2024

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Uh oh!

Conversation

quic-xuezha commented Dec 12, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Detailed description:

Pull Request Readiness Checklist

Uh oh!

asmorkalov commented Dec 12, 2024

Uh oh!

asmorkalov commented Dec 12, 2024

Uh oh!

Uh oh!

Uh oh!

asmorkalov commented Dec 17, 2024

Uh oh!

quic-xuezha commented Dec 17, 2024

Uh oh!

quic-xuezha commented Dec 17, 2024

Uh oh!

asmorkalov commented Dec 17, 2024

Uh oh!

quic-xuezha commented Dec 18, 2024

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

quic-xuezha commented Dec 12, 2024 •

edited

Loading