Edge and Region Synergy Driven Active Contour Models for Natural Image Segmentation
Satirtha Paul Shyam, C. M. A. Rahman, Pijush Kanti Roy Partho, Rahat K. Bhuiyan · 2023
Image segmentation is one of the fundamental operations of image processing and Active Contour Models (ACM) are one of the most extensively used algorithms for instant segmentation. The main benefit of ACMs is that it does not require training samples to generate reasonably accurate segmentation of image objects of complex structure, location, and size. Recently, ACMs could achieve much better results in terms of accuracy, initialization robustness, and optimization speed through the combination of edge and region-based information in energy functions. Accordingly, this work discusses, analyzes, and compares the performance of two recently reported synergy-driven ACMs in segmenting images, especially natural images. The experimental results show both models could positively augment the efficacy of the originally developed model in all the performance-assessing parameters of image segmentation.