步态识别是一种根据人行走方式的不同来进行身份识别的生物特征识别技术。本文构建了一个FPGA控制的步态识别系统,实现了对人的自动、准确、实时的识别。系统的功能为:在简单背景下,从视频序列中分离运动目标,计算其特征值,并与步态数据库数据进行逐一比对,进行鉴别。采用构建的系统对中科院自动化研究所的步态数据进行识别,达到了90%的识别率。%Gait recognition is a biometrics, which is to identify one's identity according to the different ways of walking. This new access gait recognition system based on FPGA control, involves of auto, exact, real-time identify recognition. In a simple background, the moving objects are separated from the input test video sequence, then the characteristic value is calculted, and gait database data for identification is compared. In the end, this technique is applied to CASIA gait data-set, which achieves probability of correct recognition of 90%.
展开▼