Geometric active contour detection using gradient vector flow and shape-based image segmentation

In computer vision and pattern recognition, the role of image processing for image segmentation or object boundary recognition plays a key role. In traditional snake algorithm the boundary of image is considered as parametric curve. The process of finding an object boundary has become an energy minimization process. In the project work a combined process of GVF Snake algorithm which has larger capture range and stronger convergence ability to boundary concavities than traditional snake and SUSAN approach has been implemented. The Corner points at the edge are checked first using Susan approach and then those points are marked as energy minimization points, then GVF Snake model has been used to capture object boundary after set initial snake curve. The results obtained indicate that the combined approach of SUSAN and GVF Snake algorithms based segmentation process and further building and specialized object recognition system, which is being training by segmented images obtained by this approach can improve GVF snake model´s precision to capture the boundary with sharp-angled corner as well as Object recognition system.