Graph Cuts-Based Active Contour Model with Selective Local or Global Segmentation

We are pleased to announce the publication of our latest research paper titled "Graph Cuts-Based Active Contour Model with Selective Local or Global Segmentation." This paper presents a novel approach that combines graph cut optimization with active contour models for more precise and controllable image segmentation. The implementation features an energy minimization framework where regional terms are optimized using max-flow algorithms, allowing selective segmentation through adjustable local/global constraints. The model's core functionality includes region-growing initialization, boundary term calculation using gradient information, and graph construction with capacity edges representing regional and boundary penalties. The attached source code demonstrates the complete pipeline from image preprocessing to final segmentation mask generation. We welcome researchers and practitioners to download our paper and source code to explore the technical details and implementation specifics. Thank you for your interest in our work!