Study On Targets Tracking Algorithm And Detection Processing Technology

Dynamic target tracking has been the important composing of the aviation, spaceflight, weapons guidance and control system, its precise location tracking effect can effectively improve the intelligent controlling efficiency of the system. In target tracking system, synchronized controlling of tracking control platform, detection and recognition processing of target are the core of the controlling system, they restrict the stability of system. Due to the variability of environmental and the diversity of target, the tracking aim of all kinds of dynamic targets is not the only, and there are a lot of unpredictable difficulties. Especially duo to variations of dynamic parameters about long-distance dynamic target, such as acceleration, deceleration of target, changes of the background environment, inconsistent state of the loading platform, there are a lot of difficulties in remote dynamic targets tracking system. Based on the background, real time tracking and detection method on dynamic target is studied. In the paper, the method provide data analysis gist for long-distance dynamic target,such as, information acquisition, intelligent controlling, fault diagnosis and so on. The main research and innovation work is shown as follows.Tracking controlling method and the controlling principle of target is studied, according to the relationship of a predetermined route, dynamic target and controlling platform, the geometric model between dynamic target and controlling platform is established. According to theory speed of the target by a route, the target function between displacement of target by a route and the controlling offset of turntable angle of controlling platform is deduced. Using geometric characteristics of the optical system, dynamic target image in the different time by whole route is gained, then,combined with the calculation formula, target locations is confirmed, accordingly the aim of real-time monitoring and tracking the target is achieved.In this paper, firstly,the improved Mean Shift target tracking algorithm is put forward to track the taeget, then, combined with the collected target image, the validity of the algorithm is verified,and the algorithm overcomes the shortcomings of traditional Mean Shift tracking algorithm if the target contrast comparison is low. Secondly, In order to improve efficiency of the target detection, wavelet transform is used to preprocess image, and multi-scale adaptive algorithm is used to extract edge feature. Finally, high order statistics of wavelet packet transform is used to detect target, In this method, high order statistics criterion is studied, accordingly, wavelet packet transform method is applied to dynamic target tracking.According to the characteristics of dynamic target and gained target image, dynamic target image processing technology is studied. Firstly,Butter worth low-pass filtering, median filtering, neighborhood averaging filtering, and Wiener filtering are used to preprocess infrared target image by a route, the pretreatment effect show that the median filtering can meet the demand of the system. Secondly,Technology of separation target from background in the target image is studied and maximum classes square error method is used to process dynamic target image segmentation. Secondly, the Roberts edge detection operator, Prewitt edge detection operator, Sobel edge detection operator, and Canny edge detection operator are used to recognise and extract edge of infrared target image, the results showun that Canny edge detection operator has a good single edge characteristics, can make the effect of background noise small, and is propitious to recognise and process target in complex environment. Finally, combined with the characteristics of the tracking system, based on image recognition, improved difference image method is study to process continuous image of target by aroute.Through studying the characteristics of target, based on target tracking platform, in accordance with sliding parameters of the goal, and combining with the theory flight speed of target, the relation between rotation parameters of tracking turntable and target tracking orientation is analyzed. Controlling algorithm on tracking turntable is studied, and the affecting factors of the dynamic target tracking is analyzed.Combined data transmission requirements of tracking platform, the transmission reliability and energy efficiency of wireless sensor network are analyzed, the calculation and simulation experiment show that the network coding can improve the reliability of data transmission, optimal node spacing scheme can balance the network load and reduce the energy consumption of sensor nodes, and,the optimal scheme of node spacing combining with network coding scheme can effectively improve the success rate of the transmission and reduce the energy consumption of sensor nodes. Through the acquisition sequence image processing and analysis, difference between the measured angle and theoretical control angle, and difference between measured speed and theoretical speed validate and improved algorithm theory.