【摘要】 手势识别是手语翻译中的一个组成部分,笔者提出一个使用OpenCV2实现实时识别单手数字手势方法,使用基于自适应混合高斯背景建模的背景减除法获取手部特征,然后根据手势角度判断输出的数字。实验结果表明,在日常生活场景中,该方法能识别连续动作的手势,具有较强的实时性。更多还原

【Abstract】 Gesture recognition is an integral part of sign language translation. The author proposes a method for real-time recognition of one-handed digital gestures using OpenCV2. The background subtraction method based on adaptive Gaussian background modeling is used to obtain hand features, and then the output is judged according to the gesture angle. digital. Experimental results show that in daily life scenes, this method can recognize continuous gestures and has strong real-time performance.更多还原