Region and Active Contour-Based Segmentation Technique for Medical and Weak-Edged Images

One of the key requirement in image guided surgery (IGS)/ computer aided surgery (CAS) planning is accurate segmentation of the images concerned. It is also a challenging issue for the purpose of image analysis and understanding in general, and surgical intervention involving image guided surgery (IGS) in particular. Thus, in this paper, a technique employing two-stage segmentation in which one of the stage is also an hybrid of two segmentation methods is developed for medical images in particular, and weak-edged images in general. The first stage employs hybrid of multiple-thresholding and correlation matching. The output image of the first stage was use as the input image to the second stage to generate the final output using the modified Chan-Vese level-set algorithm (MLSA). The results obtained is accurate as showed in figures 2, 3, and 4.