Marked Point Process Model for Curvilinear Structures Extraction

摘要: In this paper, we propose a new marked point process (MPP) model and the associated optimization technique to extract curvilinear structures. Given an image, compute intensity variance rotated gradient magnitude along line segment. We constrain high level shape priors of segments obtain smoothly connected configuration. The consists two steps reduce significance parameter selection in our MPP model. employ Monte Carlo sampler with delayed rejection collect hypotheses over different spaces. Then, maximize consensus among detection results reconstruct most plausible structures without estimation process. Experimental show that algorithm effectively localizes on wide range datasets.