Using Geodesic Active Contours for motion-blurred images contour detection
Gang Xu, Lei Shi · 2008
Active Contours Model (Snake) is a typical quantity dynamic contours model. Its main principle is to define a curve with energy in the area in which the researcher is interested, approaching the contours of the object through optimizing the energy function dynamically. Its biggest defect is that it can hardly handle the topological structure transformation when extracting the object contour. Geodesic Active Contours Model (GAC), whose curve evolution equation doesn’t include any parameters that have no relation with the curve’s geometric structure, is based on the theory of curve evolution and Level-Set, so GAC can automatically process the topological structure transformation when extracting object contour. This paper proposes a new method of motion-blurred images contour detection. This method, which is based on the theory of GAC and the blurred image restoration, completely detects motion-blurred images’ contour. The theory and experimental analysis show that this new method has good effects.