Research On Tracking Of Imaging Area Target

Imaging homing has been widely used in many kinds of guided weapons, the imaging area target tracking is one of key information precessing techniques. In order to satisfy the requirement of tracking, area target tracking algorithms were studied in this dissertation, include:(1)An improved image segmentation algorithm based on OTSU is designed. OTSU was implemented iteratively according to the idea of P parameter methods, which solve the problem of getting wrong threshold when the intensity changed largely in the area of target (or background). Simulation results indicate that the proposed algorithm not only is able to find target correctly and completely, but also can keep compatible with the OTSU.(2)An improved area target tracking method on the basis of image matching is designed. The strategy of one dimension global searching based on row and column projection and two dimensions local searching were combined to search the target, which not only can obtain the global optimum matching point, but decrease the computation. The selected pixels were used to compute the shape center of the target so as to locate the target in sub-pixel. The template was updated intermittently by setting the updating condition, and the computation was reduced further. Experiment results demonstrate that the presented algorithm can track target exactly and stably, and the location precision reaches sub-pixel.(3)The above algorithms were transfered to the information processor with DSP as the core. In the process of transfer, the algorithms were improved according as the characteristic of DSP and the program was optimized. The results of performance testing illustrate that the improved algorithms keep the good performance in location and meet the requirement of real time.