Graph Spectral Image Segmentation

Threshold-based, edge-based, region-based and energy-based approaches have been applied to many image processing applications successfully, for example, in medical imaging, tracking and recognition. By using the representation of graphs, morphological processing techniques can be applied to obtain many interesting image segmentation results. This chapter focuses on the concept of graph image segmentation methods. It discusses the constrained optimization model arising from the graph image segmentation problem. The chapter presents the two-class model to the case of multiple-class image segmentation. This multi-class model allows us to handle images with multiple segments. Foreground-background segmentation has wide applications in computer vision, computer graphics and medical imaging. The optimization models have been successful in segmenting single images. In image segmentation, a cost function usually consists of the two terms: the region term and the boundary term. The Mumford-Shah model is an image segmentation model with a wide range of applications in imaging sciences.