The Extraction Of Image Information Based On Optical Flow Techniques

Optical flow estimation is a low-level phase of motion recovery problem. Its output is a 2D projection of the velocity field on the image plane, which is analyzed further to get high-level motion descriptions. Optical Flow computing don't require the rigorous corresponding relationship among features of sequential images, so this approach is widely used in computer vision field including detection and dynamic analysis of moving objects. The gradient-based approach sponsored by Horn and Schunck for the optical flow estimation is one of the important kinds. The classical Horn-Schunck optical flow algorithm is discussed, where first analyzing two constraints including brightness conservation constraint and smoothness constraint, and then deducing computation formula. Finally, the results of the optical flow field using different parameters are given. As the traditional optical flow algorithm can't detect the moving objects in cases that the displacements between two continuous frames were less than one pixel, we developed a modified gradient-based optical flow algorithm. The process of gradient computing used 3D-Sobel operator. The experiment result shows that the algorithm can accurately and validly detect different objects with different velocities in static background. A method using in aircraft of passive ranging from optical flow of infrared object was proposed. The range was achieved through analyze of the movement of the imaging sensor and the optical flow of the object. A range expression at each pixel may directly calculate from pixel location and ownship motion. The range to the object is the average pixel ranges of the object area. The range can be used to enhance navigation and improve trajectory shaping.