This report summarizes the work accomplished under Phase I of the Dynamic Image Disparity Analyzer DIDA program. The DIDA program was initiated to investigate the problems associated with real-time or dynamic disparity analysis. Image disparity analysis is the determination of geometrical differences between two or more images caused by binocular parallax, camera motion, object motion, or some combination of these. The disparity field can provide important 3-D information about structure and motion which is impossible or very difficult to derive from a single image. This report outlines DIDA related efforts however, its main focus is on in-house achievements. A gradient statistics formalism was developed in order to provide a unified treatment of interest operators for selecting matchable points in different images. This formalism was also used to develop several new local operators for characterizing orthogonal and alignment aspects of local gradients. Computing complete image-to-image disparity maps in real-time is currently an unsolved problem.