复杂背景下红外小目标实时检测与跟踪-模式识别与智能系统专业论文.docx

华中科技大学硕士学位论文 华 中 科 技 大 学 硕 士 学 位 论 文 II II Abstract The target detection and tracking systems using infrared image are widely used in various fields, such as missile controlling-and-guiding, medical treatment, traffic control, etc. because they could process multi-dimension signal and are passive systems. With the developing requirements in applications, the targets are required to be detected as soon as possible. So there is much interest in studying of the detecting and tracking technology of dim small moving target from image sequence. In this dissertation, we mainly study the technology of IR small target real-time detecting and tracking in complex background.The main contribution of the dissertation is as follows: The techniques of small target detection and imaging characteristic in IR image are analyzed and studied principally. For the bright and dark target detection in IR image, an algorithm of local contrast enhancement based on lateral inhibition is proposed. Experiments results demonstrate the approach proposed to be effective in detecting small bright and dark target in complex background. Based on the analysis of target tracking techniques, the data association in multi-targets tracking is mainly studied. Combined to the imaging characteristic of IR, a data association algorithm based on multiple features adaptive fusion is proposed and implemented in order to ensure the reliability of data association. On the basis of IR imaging tracking system, an algorithm of targets detecting and tracking in complex background is researched and implemented. Based on the discussion about algorithms transfer, the algorithms switch between small and area target detecting and tracking is implemented. The realization of algorithms on IR tracking system is researched. Based on the analysis of program optimizing on DSP, some functions are optimized. Besides, a structure of space parallel with pipelining is adopted, and the algorithm is distributed on the multi-DSPs system