cv2.imread(img, -1) for IMREAD_UNCHANGED

[tudoulei]

if you train a 4 channel model, Using the detet.py code will report an error

🛠️ PR Summary

_{Made with ❤️ by Ultralytics Actions}

🌟 Summary

Improved image loading in YOLOv5 to support various image formats.

📊 Key Changes

• Modified the image reading function cv2.imread by adding the -1 flag.

🎯 Purpose & Impact

• Purpose: To enable the loading of images with their original depth and channel settings (e.g., alpha channel in PNGs).
• Impact: Users can now train models with images that have multiple channels or non-standard encodings without pre-processing them. This could lead to more versatility in training datasets and potentially richer model inputs. 🌈🖼️

[github-actions]

👋 Hello @tudoulei, thank you for submitting a 🚀 PR! To allow your work to be integrated as seamlessly as possible, we advise you to:

• ✅ Verify your PR is up-to-date with origin/master. If your PR is behind origin/master an automatic GitHub actions rebase may be attempted by including the /rebase command in a comment body, or by running the following code, replacing 'feature' with the name of your local branch:

git remote add upstream https://github.com/ultralytics/yolov5.git
git fetch upstream
git checkout feature  # <----- replace 'feature' with local branch name
git rebase upstream/develop
git push -u origin -f

• ✅ Verify all Continuous Integration (CI) checks are passing.
• ✅ Reduce changes to the absolute minimum required for your bug fix or feature addition. "It is not daily increase but daily decrease, hack away the unessential. The closer to the source, the less wastage there is." -Bruce Lee

[glenn-jocher]

@tudoulei thanks for the PR! It appears you change updates the second argument of cv2.imread() which specifies the ImreadModes, -1 being 'unchanged'.
https://vovkos.github.io/doxyrest-showcase/opencv/sphinx_rtd_theme/enum_cv_ImreadModes.html

Can you provide an example image that is causing an issue with the default code please? Thanks!

[glenn-jocher]

Profiling results are acceptable.

%timeit cv2.imread(path, -1)
5.26 ms ± 54.1 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)

%timeit cv2.imread(path)
5.27 ms ± 63.5 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)

[glenn-jocher]

@tudoulei thank you for your contributions to YOLOv5 🚀!