Research And Implementation On Variable Scale Target Detection And Tracking Of Passive Millimeter Wave Imaging

Passive millimeter wave(PMMW) image can be obtained by detecting the radiation energy difference of scene in millimeter-wave band. Relative to the optical, infrared and microwave imaging technology, it has its unique characteristics and advantages. Passive millimeter wave image variable scale target detection and tracking technology extracting and tracking the target from the background using the radiation differences of objects and providing more valid information about the target of interest for different applications. It is a very important issue in the Millimeter-wave field and areas of accessing informations.The thesis, surrounding the PMMW imaging system application requirements of the scene surveillance and security applications, focuses on the detection and tracking algorithm in the near-field environment of the human body to carry hidden metal objects. The main contents of the thesis are as follows.1. Focus on the commonly used theories and methods of the PMMW imaging for target detection and tracking, Study and analyzed the theoretical basis for this essay suitable algorithms. Analysis the characteristics of the hidden metal objects in PMMW images, compare their advantages and disadvantages, research to establish the overall framework for detection and tracking algorithms.2. Aimed at the displacement and scale changes brought occult metal target detection and tracking difficulties, Improve the detection algorithm according to the NP criteria to determining the basis binarization threshold and establish the space of threshold, voting to detect the target object, and the tracking algorithm based on the Kalman filter to estimate the target location and scales. Research and analyzes the performance of the algorithm, proved the algorithm can achieve higher accuracy in detection and tracking of metal targets.3. In view of the parameters of the PMMW project in signal processing platform, finished the selection and the interface debugging of DSP.4. Completed the migration of the algorithms, and the DSP program optimization which based on the C64x+ DSP core features and various optimization strategies. Ultimately results showed good detection and tracking performance of the algorithm.The above work has been successful verified in Matlab software platform and DSP hardware platform, proof that the algorithms can achieve good detection and tracking efficiency and hardware real-time and they can be used to solve the problems of detection and tracking in the near-field environment of the human body to carry hidden metal objects.