Enhancing Video Object Segmentation Results Through Biologically Inspired Postprocessing
Dubravko R. Culibrk, Vladimir Radenkovic, Daniel Socek · 2007
Object segmentation from a video stream is an essentialtask in video processing and forms the foundation of sceneunderstanding, object-based video encoding (e.g. MPEG4), andvarious surveillance and 2D-to-pseudo-3D conversion applications. Many segmentation approaches are pixel-based and sufferfrom noise in the segmentation results, due to the fact that theseapproaches do not exploit spatial information. Morphological post processing is typically used to enhance the segmentation results. Here, an alternative post-processing approach is presented. The proposed approach is inspired by well-known aspects of the primate visual system function. The approach is particularly suitable for use with the probabilistic segmentation algorithms and allows for efficient neural-network-based implementation.