fix padding for rectangular inference

[menggui1993]

Image size will be padded to minimum multiple of 32 for rectangular inference.

🛠️ PR Summary

_{Made with ❤️ by Ultralytics Actions}

🌟 Summary

Improved image padding granularity in YOLOv5 pre-processing.

📊 Key Changes

• Modified the padding calculation in the letterbox function within datasets.py.
• Padding values are now a multiple of 32 instead of 64.

🎯 Purpose & Impact

• 📈 This change enhances the flexibility in image processing by allowing finer-grained padding, which can improve model accuracy as it provides more precise scaling options.
• 🔍 Users may notice improved detection performance, especially with small objects, due to less distortion during image resizing.
• 🚀 Models could potentially train and infer slightly faster as a result of the reduced padding in some cases.

[github-actions]

Hello @menggui1993, 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 update by running the following, 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/master
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]

@menggui1993 thank you for your contributions.