This book investigates the interconnected components of image segmentation and object/scene recognition, focusing on Deformable Contour Methods (DCMs) for medical image segmentation. It presents a comparative study of eight major snakes and level set methods, evaluating their strengths and limitations through qualitative and quantitative measures. The findings from this medical segmentation analysis are then extended to other domains, demonstrating how DCM results can be recognized for further image analysis. Additionally, a skeleton-based graph matching algorithm is introduced for challenging segmentation applications, enhancing accuracy through feedback into the segmentation process.

The book provides a detailed comparative analysis of Deformable Contour Methods (DCMs) for image segmentation, with a particular focus on their application in the medical field. It aims to offer insights into the strengths and limitations of various techniques, drawing lessons that can be applied across different segmentation domains. The inclusion of a skeleton-based graph matching algorithm suggests a focus on practical solutions for complex segmentation challenges. The work appears to be a technical review and methodological proposal for researchers and practitioners in image processing and computer vision.