【摘要】 为解决生物识别中,手形特征维度高、计算量大、易受图像背景及手掌位移干扰等问题,提出一种基于压缩相对论廓特征点的手形识别方法.建立了手形轮廓相对于手形轮廓质心的坐标系,通过等间隔取点法寻找综合计算性能和准确率最好的手形特征点数量,以此数量等间隔提取手形轮廓,并用压缩特征点坐标作为手形识别特征.采用支持向量机作为分类器对不同手形特征进行识别,准确率达98.25%.更多还原

【Abstract】 In order to solve the problem that in biometrics, due to the high dimensionality of hand shope features and a large amount of cakulation, it is susceptible to interference from the image background and palm displacement, a hand shape recognition method based on compressed relative contour feature points was proposed. The coordinate system of the hand contour relative to the centroid of the hand contour was established, the number of hand shape feature points with the best comprehensive calculation performance and accuracy was found through the method of taking points at equal intervals, the hand contours were extracted from this number at equal intervals, and the coordinates of the compressed feature points were used as hand shape recognition features. Using support vector machine as a classifier to recognize different hand shape features, the accuracy rate is 98.25%.更多还原