Image segmentation method

摘要

The image segmentation method involves dividing each image into regions, with the adjacent points within an image region having similar brightness and/or colour values. A defined region within one image (In) is located within the following image (In+1), with an estimation of the movement of the image regions. The movement vector for each region is used to provide the new position of the region, with a matching of the shifted regions to the image content, and provision of a new regions for the image points which do not match the shifted regions.