Research On The Detection Of Small And Dim Target In Infrared Image Based On Morphology Filtering

The technology of small target detection and track system in infrared image sequences is the core-technology in the system of infrared detection and tracking. And the detection of small target is an more important project recent years. The current development trend is the new method with little operations, high performance, and which can be achieved by hardware in real time.Combined with many detection algorithms given by the former, the applications of the Mathematical morphology method in small target detection is mainly discussed in this paper. It begins to form the discussion the knowledge of the two value morphology and the grey degree morphology and the formation element , which are the base of Mathematical morphology method, then a method to detect the small target in IR by Top-Hat Transform of the Mathematical morphology be advanced to the method through the DSP and FPGA device. First the noise and the air background of original image is suppressed by Top-Hat Transform of the Mathematical morphology, and the target is segmented by open process, then the little target signal in IR could be detected with Adaptive threshold. The results show this method can significantly enhance SNR of target, and it is adaptable to real-time image processing.Considering the application in engineering, this paper also gives the particular design idea and flow of this method's realization in DSP and FPGA device. Because of the all-around consider about the effective utilize of the resource in this system, the small infrared target can be detected in real-time by this method with this device system. Inthe end, the prospect of the project is as well presented.