基于手指角度特征的静态手势识别算法

摘要

This paper puts forward a static gesture recognition algorithm based on finger angle characteristics .With the fingertips attachment the authors constitute the gesture framework by the line from finger to the center of palm , and calculate the angle between your fingers.To view the number and size of classification , the gestures are defined as one , two, three, four, five, six, seven, eight, nine.The algorithm is not affected by gestures to the direction and scale , and only by judging the fingers point size to recognize hand gestures.In experiments, 900 gesture image recognition are analyzed , the accuracy rate is 96.8%, the average is less than 0.05 sec-onds, which indicate that the algorithm is of high accuracy , good real-time performance.%提出一种基于手指角度特征的静态手势识别算法。以指尖到手掌中心的连线构成手势骨架，计算手指间的角度；以角度的大小和指间数量进行分类，把手势定义为一、二、三、四、五、六、七、八、九等9种。该算法不受手势的方向和尺度的影响，仅通过判断手指间的角度大小来识别手势。对900幅手势图进行分析识别的实验结果表明：该算法正确率达96．8％，准确性高；平均用时不超过0．05 s，实时性好。