The non⁃overlapping regions of different monitoring cameras,calledblind area of the camera,pose challenges to matching the humansamong the cameras.This paper presents a method of human matching based on block features fusionfor blind areas of multiple cameras.First the discriminative training component model trained by HOG features is used to detect humans within the field of view of a single camera,then COLOR,SIFT andLBP square block features are extracted from the Hue component of the detected target to find the best fusion features through experiments for matching the humans. Experiments show the proposed matching method with the best fusion features has higher speed than the method in the literature.%各相机间通常存在监控视域以外的非重叠区域，即相机视域盲区，它的存在对多相机多行人匹配带来了挑战性工作。提出了一种基于分块特征组合的多摄像机视域盲区行人匹配的方法。在单摄像机视域中首先通过 HOG特征训练而成的判别训练部件模型对行人进行检测，在其Hue分量上分别提取COLOR、SIFT 和 LBP 分块特征，从而通过实验找出最佳组合特征实现行人的匹配。提出的最佳组合特征和已有文献中的匹配方法进行比较，结果表明匹配速率有很大的提高。
展开▼
