立体视觉Gaussian Super-pixel Based Fast Image Segmentation Using Graph Cuts.-Gaussian Super-pixel Based Fast Image Segmentation Using Graph Cuts.

This paper addresses the problem of segmenting an image into regions. We de¯ ne a predicate for measuring the evidence for a boundary between two regions using a graph-based representation of the image. We then develop an e±cient segmentation

In the proposed algorithm, these two issues are solved by a novel predicate, which is defined by the sequential probability ratio test (SPRT) and the minimal cost criterion. Starting from an over-segmented image, neighboring regions are progressivel