Infrared Small Target Detection And Tracking Algorithm

Infrared imaging system is suitable to be applied in the field of detecting, tracking and guidance. It has become a critical technique in the modern military navigation system for its advantages including the special passive working way, concealment, anti-jamming and high precision tracking performance. At present, the application of infrared imaging technology in target detection and tracking has been a widespread concern at home and abroad. It is a very important issue that has been looking for a real-time effective improvement and innovation based on the existing detection and tracking algorithm. The study is looking for certain new applied arithmetic of small targets detection tracking in infrared background. The purpose of the research is to make tracking more accurate and precise based on the lower time-consuming operation from the angle of restraining.In recent years, many native scholars and foreign experts have been making a deep research and put forward many small targets detection and tracking algorithm under complex infrared background. The gray value of infrared image always changes remarkably in complex background, where target is easily submerged and noise can not be detected. There is not a detection and tracking algorithm which can completely solve all the problems in present stage. Therefore, it is a major task to make a deep research in IR target detection and tracking. This dissertation analyzes deeply the achievements of other researchers for reference, and then proposes three improved target detection and tracking algorithms based on complex background as follows:(1) A morphology-based algorithm is presented to detect infrared weak point target. The new algorithm preprocesses each frame image of an infrared video sequences with mean filtering and opening top-hat filtering, and then the moving small IR target is extracted by means of statistical segmentation algorithms based on the average and square errors. The experimental results show that the presented method has a better detection and segmentation performance, and the tracking accuracy is improved effectively.(2) Based on classification background prediction and image block technology, a small IR target detection algorithm is presented to detect small infrared targets by use of the principle of the background prediction and image block according to the feature of small IR target in complex background. By the use of edge detection, maximum mean algorithm and partial similar methods, this approach can obtain a relatively accurate background. With the adoption of the statistical segmentation method, the proposed method can obtain a more accurate position of target from the residual images. In the process, the image block technique is used to improve the efficiency of the algorithm. Finally, the computer simulations show that the presented algorithm has obvious advantages in the detection accuracy, robustness, and calculation efficiency.(3) With the analysis and discussion of the kalman filter, two small targets tracking algorithms are proposed on the basis of classification background prediction as well as kalman estimation:sub frame local search algorithm and global threshold algorithm. The two improved algorithms are all based on the classification background prediction algorithm, the former algorithm sets up a sub frame box in the center of KF forecasting results, and then finishes classification background prediction and correction in this sub frame. The latter algorithm makes a threshold judge after KF forecasting and classification background prediction, and then corrects the discrimination results. Finally, through a series of the simulation results, the performance of the new algorithms is verified in infrared small target tracking. The experimental results show that the new algorithms will get an accurate and effective tracking effect in complex background.