Abstract
手指康复运动具有高移动精度与高控制分辨率的需求,加上因人而异的数据多样性,以及在运动过程中的模糊信号识别等问题,实时准确的手指姿态识别能够大大提高手指的康复效果。针对这一问题,本文提出了一种简易、便携的手指姿态识别系统设计。使用多分类支持向量机对手部运动进行分析和识别,大量训练者所采集到的数据分为离线训练集和在线测试集,经过大量的离线训练与在线测试,结果表明,多分类支持向量机在分类和识别过程中具有高效性和实用性,并且极大可能有助于手指的康复过程。 With high precision and high resolution of the mobile control demand, plus the data, it differs from man to man. Diversity, and in the process of movement of fuzzy signal recognition, real-time and accurate finger gesture recognition can greatly improve the effect of rehabilitation of finger. To solve this problem, this paper presents a simple, portable finger gesture recognition system design. The use of multi class support vector machine hand motion analysis and recognition, a large number of training data collected is divided into offline and online training set test set; after the test, a large number of online and offline training results show that the multi class support vector machine is efficient and practical in classification and recognition in the process of the rehabilitation process and which may contribute to the finger.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have