[′pik·chər ‚seg·mən′tā·shən] (computer science) The division of a complex picture into parts corresponding to regions or objects, so that the picture can then be described in terms of the parts, their properties, and their spatial relationships. Also known as scene analysis; segmentation.