基于HSV模型和特征点匹配的行人重识别算法

提出了一种基于HSV模型和特征点匹 配相结合的行 人重识别算法。首先根据改进的HSV空间颜色量化策略,比对两幅行人图像的 躯干和腿部主颜色是否一致,以快速确定备选目标；然后对备选目标,利用环形Gabor滤波器组生成多尺度图 像,再利用改进的FAST算法和BRIEF算法对多尺度图像进行特征点提取与描述,最后利用暴 力算法和随机抽样一 致性算法进行特征点匹配和提纯,以达到较好的匹配效果。实验结果表明,本文提出的行人 重识别算法具有较高的识别准确率,识别速度达到12frames。

Person re-identification algorithm based on HSV model a nd keypoints matching

In the application of video content an alyses and multimedia retrieval,person re-identification is a critical technique,which has great significance in the field of criminal inves tigation.This paper proposes a person re-identification algorithm based on HSV model and keyp oints matching.It first utilizes HSV model to pre-test pedestrian images and quickly rule out images whose main colors are different from the target,and then tests remaining images by matching keypoints.Then,the method of pre-testing pedestrian images compa res main colors of the torso and legs of two images according to an improved color quantization strategy of H SV space,which achieves better pre-testing effects and raises recognition speed of the algorithm.The method of keypoints matching first takes advantage of circularly symmetrical Gabor filters to generate multi-scale images,then extracts and des cribes keypoints by improved FAST and BRIEF algorithms,matches and purifies keypoints by Brute force algorithm and Ra ndom sample consensus algorithm, which achieves better matching effects than SIFT,SURF and ORB algorithms.For tes ting the validity of the proposed algorithm,we establish a pedestrian image library including 600images.Ex perimental results show that the proposed algorithm can identify pedestrian targets accurately by the speed of 12frames per second.