Design And Implement Of Preprocessing Algorithm For Dim And Small Target In Infrared Image

Infrared imaging guidance is a hot topic of modern high precision guided weapons, because of its high sensitivity, strong anti-jamming capability, attacking target around-the-clock and in all directions. The dim and small target detection in infrared images is one of the key techniques in military weapon system.To the question of infrared imaging guidance system, this paper has deeply studied and discussed some kinds of background suppression methods and target detection algorithms and chosen a series of methods which were easy for hardware implement. At last, these methods have been implemented on infrared image processing hardware platform.According to the characteristic of infrared dim and small target, this paper has researched temporal-spatial fusion filtering algorithm, improved median difference filter, researched small structure element eroding algorithm, applied genetic algorithm to image segmentation and introduced infrared image processing system design scheme and hardware implement in detail.The result of experiment proves that this method avoids complicated computing and huge computational requirement. It reduces computational requirement, accelerates computing speed and achieves a high availability and reliability in dim and small targets detection. The results on the hardware platform shows that the hardware runs stably and the method can detect in real-time, which satisfies the design requirements.