An image analysis method has been proposed to segment color images. The method is capable of segmenting color images based on watershed transform and scale-space analysis of the images. We analyze hierarchical, geometric properties and the hue uniformity measure of the tree of scale-space and show the experimental results for several biological and medical images. The experimental results show that the proposed method provides effective segmentation. Moreover, this approach preserves the object contour, an important feature because of the performance of measure and other chemical processes are greatly influenced by their morphological character.