The field of image processing addresses handling and analysis of images for many purposes using a large number of techniques and methods. The applications of image processing range from enhancement of the visibility of cer- tain organs in medical images to object recognition for handling by industrial robots and face recognition for identification at airports, but also searching for images in image data bases. The methods applied range from low-level approaches like boundary detection and colour based segmentation to advanced object detection using statistical geometric models. Often several techniques must be combined to obtain a desired result, e.g. first low-level feature extraction, next clustering into regions, extraction of shape parameters and finally object recognition. Whereas image processing basically includes all thinkable operations on images its sub-field image analysis addresses the extraction of certain information from images and aims to generate a description of (part of) the image or objects present in the image. In this chapter the stress will be on image analysis rather than on image processing in general and on static images rather than on image sequences. Examples of generating descriptions of images are recognition of a face in an image, counting the number of a certain type of cell in an image and labelling the different organs in a CT image of the chest (heart, lungs, ribs). Thus, given an image, image analysis aims to generate a description. On the other hand, the common task of image processing in multimedia data base applications is to find images based on a description, where the description can range from an abstract description to an example of an object in the form of another image. Of course, both viewpoints are closely related, because in order to find an image based on a description, one must also be able to generate a description from an image.