The Research For Detecting And Tracking Of Moving Targets

With the development and application of the digital video technology, a new research for detecting and tracking of moving targets is produced. For example, the tracking and recognition of moving targets in artificial intelligence systems; the monitoring of moving vehicles in intelligent transportation systems; the detecting and tracking of moving targets in security systems such as banks , warehouse, etc; the tracking and recognition of military targets in national defense security systems; the application in biomedicine domain; etc.This thesis includes the construction of the real time video sampling system, the detecting and extracting and tracking of moving targets in static scene.In the processing of constructing the real time video sampling system, the constitution of hardware and software of the video sampling system is introduced, and each functional unit of hardware platform is introduced. The image processing based on high-performance DSP TMS320C6713 is produced and the problem of video Image Acquisition.In the aspect of the research for the extracting and detecting of moving targets, several common algorithms are first introduced, then the algorithms of extracting moving targets is studied by experiments and the algorithms of image preprocessing, image segmentation and image matching are discussed.In the aspect of the research for the tracking of moving targets, several common algorithms of target feature and matching are first introduced, then the design scheme and algorithm are simulated by Matlab7.0, and in simple indoor scene the results of moving targets shaded by obstacles are discussed and the superiority of the relative tracking algorithm is verified by experiments. The design gets satisfied results.The system of the extracting and detecting of moving targets become a study hotspot present an future, but the comprehensive performance of the system such as real time, the accuracy of the detecting and recognition of moving targets, the stability of tracking, etc, is still improved.