摘要
提出了一种在距离空间内采用Euclidean距离计算的手势识别算法。手势图像经过边缘检测后,对边缘图像实施Eu-clidean距离变换(EDT),在距离变换空间内计算距离映射图与样本之间的Euclidean距离。最后,对基于单目视觉的30个手指语字母手势进行识别,最好识别率达93.33%。
A Hand Gesture recognition algorithm is presented, which based on Euclidean distance, computed in the space of Euclidean. After the edge examing, Euclidean Distance Transform (EDT) is implemented, calculates the minimum Euclidean distance between the images, Finally, the recognition of 30 letter gestures is performed by computing distance. The best recognition rate of 93.33% is achieved.
出处
《微计算机信息》
北大核心
2007年第25期265-266,191,共3页
Control & Automation
基金
陕西省自然科学研究计划(2005F50)