Automatic segmentation of video object plane based on object tracking and matching

Abstract

This paper presents a new approach to the extraction of a video object and segmentation of the video object plane from a video sequence. The method is based on a morphological motion filter using the connected operator and the proposed new filtering criterion. The morphological motion filter aims to detect motion which is distinct from that of the background, and thereby locates independently moving physical objects in the scenes. Then the object track-matcher using an active contour model is presented to track and match the object against the subsequent frames. From a series of binary contours we can extract the moving object. Experiments show that our algorithm can extract objects from moving backgrounds efficiently.