【摘要】 在低信噪比条件下，要从图象中以视频速率检测出运动目标存在不少困难。提出一种用帧间邻域高位比较法检测当前帧的运动目标的方法。这种方法包括选择处理窗口对当前象素及其邻域象素编码、存储和比较。建立了该算法的硬件实现的实时系统，分析了工作时序，较好地解决了运动检测算法中出现的抗噪度与灵敏度之间的矛盾、抗噪度与实时性之间的矛盾。这种检测以视频速率实时工作，可以用于电视侦察和监视系统中，也可用于光电成象的图象去噪以及扫描速率变换中。更多还原

【Abstract】 Many difficulties are existed when the moving objects will be detected from the image with video rate under the conditions of low SNR .A detection method for the moving object within the present frame with pixels comparison between the neighboring frames is proposed in the paper.This method consist of the steps of selecting processing window,encoding the pixels of the present frame and its neighborhood,storing and comparison.The real time system implemented with the hardware of the algorithm is constructed and the time sequence is analyzed.The contradictions exised among the antinoise ability,sensitivity and real time problem in the moving detection algorithm have been perfectly solved.The detection method can be operated in video rate.It not only can be applied to TV reconnaissance and monitoring system,but also can be applied to eliminating noise in photoelectric imaging and scanning rate transforming.更多还原