An Integrated Approach to Image Sequence Segmentation

Semantic object segmentation is an important step for object based coding, content based access and manipulations. We propose a segmentation scheme for image sequences which provides initial region information for the semantic object representation of those applications. Our objective is to develop a segmentation method which has hardware friendly architecture, and incorporates static and dynamic features simultaneously in one scheme. In the initial stage, a multiple feature space consisting of luminance (chrominance), motion, and texture features is transformed to one dimensional label space by using Self Organizing Feature Maps (SOFM) neural networks. The next stage is an edge fusion in which edge information is incorporated into the neural network outputs to generate more precisely located boundaries of the segmentation. The segmentation results of both gray level image sequences and color image sequences are evaluated using evaluation metrics. The results show the validity of the proposed scheme.