An automated approach to the initialization of the snakes algorithm for the detection of swimbladder regions in x-ray image data

摘要: Snakes have been widely used for object tracking, shape detection, and segmentation of an area interest within image data. A snakes algorithm uses energy minimization approach to deform initial boundary or curve so that it traces along the contour a in image. However, major disadvantage is requires users draw object, which not feasible when large number images need be processed user-introduced bias selection may influence accurate detection objects. This paper reports on automatic region utilizes segmentation. It specifically combines multiple processing screening techniques build pipeline steps produces initialization algorithm. The has evaluated X-ray striped burrfish detect swimbladder as interest. Results from fully automated are compared against ground truth values semi-automated results.