Image segmentation based on hierarchical mapping

An efficient image segmentation technique is presented; it combines an image segmentation algorithm with a pyramidal approach to form a scale space representation. First, the coarsest level of the pyramidal image is quantized to coarse color space. Then, segmentation is achieved using the JSEG algorithm followed by region merging for further refinement. Finally, hierarchical mapping is performed to determine region boundaries in a coarse-to-fine manner using combined global and local features, until the final segmentation is accomplished. This multiresolution approach not only offers a significant reduction in computational cost, but also helps reduce the over-segmentation problem of traditional region growing and watershed techniques. Experimental results show good segmentation performance over a variety of images, and also great reduction in the amount of processing time.