Due to the increasing demand and offer of tile technology, the next generation of the image file formats will be more likely to store and retrieve images based oil their semantic content. Thus, an image should be segmented into "meaningful' regions, each of which corresponds to all object and/or background. In this study., we propose a scheme for multi-level image segmentation, based oil a simple descriptor, called "the closest color in the same neighborhood". Tile proposed scheme generates a stack of images without using arty segmentation threshold. The stack of images is hierarchically ordered in a uniformity tree. The uniformity tree is, then, associated with a semantic tree, which is built by tile user for content based representation. The experiments indicate superior results for retrieving images, which consist of fen objects and a background.