摘要
Radon变换把图像从坐标空间映射到Radon空间,因其可以保存频率信息而被应用在步态识别算法中。主要从频率角度入手,着力提高基于Radon变换的步态识别算法的识别正确率,提出了基于时间保持能量图的Radon变换步态识别算法。传统的步态能量图是对步态周期中经过归一化的人体轮廓图求算术平均而得到的步态特征表示,最近提出的时间保持能量图在保持步态能量图的优点的基础上,保留了步态序列的时间信息,在改进的步态周期检测算法的基础上,提出将时间保持能量图和Radon变换结合到一起的步态识别算法。也对结合不同数据空间的特征如频率、形状等做了初步探讨。
Radon transform is an important method in gait recognition which captures the frequency information of an im- age by projecting the 2D image from I（x,y） space to Radon space.A Radon transform and chrono-gait image combined gait recognition algorithm is raised.Traditionally gait sequences are handled in a video with Gait Energy Image（GEI） which is a compact representation of gait features by averaging aligned and unified gait silhouettes.However it also loses temporal infor- mation.Chrono-gait images solve this shortcoming of GEI while sharing all the advantages of GEI.This paper effectively com- bines Radon transform with chrono-gait image to improve the performance.Experiments indicate the promising results of the proposed approach.The proposed method achieves high recognition accuracy. Key words： frequency; Radon transform; Gait Energy Image （GEI） ; chrono-gait image; gait recognition
出处
《计算机工程与应用》
CSCD
北大核心
2011年第29期164-167,共4页
Computer Engineering and Applications
基金
国家自然科学基金(the National Natural Science Foundation of China under Grant No.60975044)
上海市科学技术委员会基金(No.09DZ2272800
No.10DZ2272800)
