Skin color detection using histogram models

[baixi-cs]

Implement color-based skin detection using 2D histogram backprojection across RGB, N-RGB, and HSI color spaces to fulfill the assignment requirements.

This PR introduces a complete pipeline for flesh tone detection, including:

• Training data collection: Optional web crawling and skin pixel extraction using either provided masks or a heuristic.
• Histogram model building: 2D histograms are constructed in RGB (R-G), N-RGB (nr-ng), and HSI (H-S) color spaces.
• Segmentation: Applying histogram backprojection to segment skin regions in test images, with options for thresholding and morphological post-processing.
• Evaluation: Tools to visualize and compare the performance of different color spaces, including an auto-selection mechanism based on IoU if ground truth masks are available.

---

 

[cursor]

Cursor Agent can help with this pull request. Just @cursor in comments and I'll start working on changes in this branch.
_{Learn more about Cursor Agents}