Why Building wheel for opencv-python when using "pip install opencv-python"
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 5.4k
- Forks
- 1k
- Avg merge
- 22h 17m
- Merged PRs (30d)
- 3
Description
System Information
OS: MacOS
opencv: Any version
Detailed description
Hello, Why does Building wheel for opencv-python work when I use "pip install opencv-python"?
Now whether I install the new environment using anaconda or install it again after uninstall opencv in the original environment, I get the steps shown below! This is completely different from my previous pip install, which didn't build before. But I don't have this problem when installing via the ".whl" file. I'm macOS and my first question is: what should I change to make my pip install opencv-python as easy and fast as before without compiling from scratch. It's too slow to compile this way。
Meanwhile pip install opencv-python, since it now has to be compiled from scratch, in addition to being a particularly slow installer has a problem that I am very puzzled by. This problem exists both when I compile opencv myself via cmake and when I make opencv-contrib available for C++. Both my own compilation of opencv (for C++) and pip's rebuild of opencv-python have no way to read png files!
C++ builds with cmake and this problem occurs whether I tick "WITH_PNG" or not:
It's worth mentioning that my system's libpng version itself is 1.6.39, so why does opencv, which is compiled for C++ with and without "WITH_PNG" checked, have this problem when reading png?
Also the previous pip install of opencv was able to read png images correctly with python's opencv since it wasn't compiled from scratch, but now since the install is going to be rebuilt, I'm getting this problem when reading opencv through python.
So my two questions in a nutshell are (1) how do I restore the normal flow of my opencv-python when installed using pip without having to compile from scratch?
(2) How to fix the problem that my own compiled opencv cannot read png images
Looking forward to your answers and help, thank you very much!
Steps to reproduce
Issue submission checklist
- I report the issue, it's not a question
- I checked the problem with documentation, FAQ, open issues, forum.opencv.org, Stack Overflow, etc and have not found any solution
- I updated to the latest OpenCV version and the issue is still there
- There is reproducer code and related data files (videos, images, onnx, etc)
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
No files, tests, or reproducible code are provided. Start by reproducing pip install opencv-python on macOS and checking whether a compatible wheel is selected, then inspect the CMake build's PNG support with and without WITH_PNG. Done means documenting the cause of the source build and restoring PNG reading in the relevant installations.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- opencv, python
- Domain
- build-system, computer-vision
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 20/100