Moving Target Recognition And Tracking Techniques Based On Infrared Scanning Image

With the developing of infrared technology, laser technology, and millimeter wave technology, their associated automatic target recognition applications are more and more. Infrared equipment has good concealment, strong anti-interference ability, far operating range and fast search speed. It has no mirror echo, can work day and night. These features make it more and more used in military or civilian fields.The target of infrared image recognition and tracking system composed by optical receiving system, infrared detector, preamplifier and the video amplifier circuit, scan converter, image signal processor and other major components. The core of the system is the image processor, it used to process the image signal, includes the preprocess of image, target detection and recognition, tracking modules.This paper aims to realize the moving target recognition and tracking system based on Infrared scanning image, that the image processor functions of the target of infrared image recognition for infrared tracking system. The contents of the main research includes:First, image pretreatment. Making image filtering, gray level transformation, image enhancement and other algorithms to the collected infrared images to reduce image noise and clutter, improve the image SNR; highlight objectives and inhibition background, prepared for the follow-up. Through the analysis and comparison of the common image preprocessing methods that include spatial filtering methods and frequency filtering method. According to the actual situation, choose a combination method of Butterworth filter and Median filter as the image preprocessing algorithm.Then, target Recognition. By choosing a suitable image segmentation algorithm to segment the target, extract the object feature vector including its rectangular degree, region moment, normalized moment of inertia, etc., make further target recognition and improve the recognition rate. Through the analysis and comparison of the three kinds of image segmentation algorithms (background subtraction algorithm, segmentation algorithm based on the target local energy, Otsu algorithm), according to the actual situation, selected the segmentation algorithm based on the target local energy, and get better results.Finally, target tracking. Select the appropriate target tracking algorithm to determine the position of the moving target, Simulated target trajectory, Identified in the coordinates. Using Kalman tracking algorithm to track the target, get the target specific position, and Converted to the corresponding display coordinates. Using GDI in VC++to establish Radar Scanning polar coordinate system, show the target trajectory in the coordinate system.