Infrared Moving Target In Complex Background Conditions, Real-time Detection And Tracking Technology,

IR moving target detection and tracking under complex environment is a key problem in IR alarm system. A lot of low altitude objects and the characteristics of point target make the detection and tracking even harder. Now there are many detection algorithms, such as nerve-network training, wavelet transform. However these algorithms couldn't reach real-time request. Tracking algorithms include centroid tracking and correlative tracking etc. Different tracking algorithms are adapted in different process for ensuring steady effect. Moreover, the solution to sheltered target is required to be considered.In this dissertation, a set of practical IR moving target detection and tracking algorithms fit for IR alarm system were presented. Firstly, to extract the movement information by decreasing two adjacent images. Secondly, to perform two-value segmentation, two-value filter. Centroid tracking and correlative tracking were respectively used in the process of point target and face target. These algorithms were simulated on the computer and the improvement of image quality was satisfying.