【作者】 王蕾； 宋文忠； 吴剑锋；

【Author】 WANG Lei , SONG Wen-zhong ,WU Jian-feng Institute of Automation, Southeast University, Nanjing 210096, China Department of Instrument Science & Engineering, Southeast University, Nanjing 210096, China

【机构】 东南大学自动化所； 东南大学仪器科学与工程系；

【摘要】 在视频跟踪系统中,实时、准确地完成对目标的自动识别是实现准确跟踪的前提。本文在分析了现有的目标识别方法优缺点的基础上,提出了一种灰度与特征匹配相结合的识别方法,并从实时性等方面对算法进行了改进,得到了一种误识别率低、准确性高且能用于实时目标跟踪系统中的目标识别方法。最后通过半实物仿真实验验证了算法的有效性。更多还原

【Abstract】 In the video tracking system, automatic target recognizing accurately in real time is the base of automatic target tracking. Advantages and disadvantages of existing target recognition algorithms are analyzed in this paper. And then a Gray-Feather matching algorithm is put forward. According to some improvements in respect of real-time character, a good target recognition algorithm with low error probability in recognizing and high veracity is obtained. This algorithm can be applied to real-time target tracking systems, which has been validated in a half physical simulation platform of automatic target tracking.更多还原