基于最大稳定极值区域的机场目标识别

摘要

针对机场背景下动态变化的场景,采用一种基于最大稳定极值区域(MSER)的算法对目标进行匹配和识别.通过采用最大稳定极值区域的方法将区域结构相似性和特征空间结合起来,比较好地实现了机场复杂背景下的特定目标的匹配和识别,提高了目标匹配识别的稳定性,并在相同试验样本下采用基于SIFT的算法进行匹配识别,将其与基于最大稳定极值区域(MSER)的算法进行比较.实验结果证明了这种方法在机场目标识别方面的有效性.%Recognition at the scene of dynamic changes of airport background,a method of object recognition based on Maximally Stable Extreme Region is presented in this paper. The method of MSER,combine the contour of the regions of the objects and the local structures of these regions in its feature space,has better property in the recognition of airport object. It enhance the stability of object recognition. This paper also compare the MSER with SIFT at the same experiment sample. The experiment results testify that the method is practicable in airport object recognition.