Research On Helmet Tracking Algorithm Based On Video Image Processing

With the development of aircraft's maneuver capability and worse of the combat environment, it is necessary for the pilot owning higher ability of Situation Awareness (SA). Apart from these, with development of abaxial missile, Head Up Display (HUD) has not met the requirement of modern air fight for the narrow Field-of-view. Hence, recently, many countries have realized the important of the Helmet Mounted Sight/Display (HMS/HMD) and equipped them for some forces. It has been an import part of Aviation Fire Control System in many countries.At present, China has not a mature HMD applied. One of the main problems is that the weight of HMD is too much. Therefore it is absolutely develop more advanced target tracking technique and automatic aim system. It was one of the new urgent issues to solve. From this research hot point, here provides a further research work in the field of helmet tracking. Some aspects have been shown in the dissertation as follows:It clarifies the concept of HMD and its working theory, and helmet tracker acts as an important role in the system. It discussed the current helmet tracking technology, analysis the current shortcomings during the process of target tracking. In order to solving this problem, it presents a new method to detection the location of the helmet based on video image processing. This method can rapidly and accurately aim target and largely decrease the weight of helmet.It demonstrates the theory about image processing, image pre-processing, edge detection, image registration and so on. The system algorithm based on image registration is proposed in this dissertation.According to the characteristics of the noise and demand of image processing, it discussed different local filtering algorithms and compares the computation results. Especially the algorithm of edge detection is analyzed. From the experimental results of different edge operators, this dissertation decides the fit algorithm-Canny edge operators. The experiment shows good results of the algorithm. Then feature was extracted, and calculated parameters of helmet position.Three projection models were discussed in details. According to the perspective projection model, it proposed an algorithm to restore spatial coordinates.As above mention algorithms, it included image filtering, denoising, edge detection, parameter calculation of helmet position. Visual C #.Net programmed the whole system and showed good results. Finally, it gives error analysis on the basis of the experimental results. Experimental result shows the method of helmet tracking based on video image processing feasible and effective.