国产欧美精品一区二区,中文字幕专区在线亚洲,国产精品美女网站在线观看,艾秋果冻传媒2021精品,在线免费一区二区,久久久久久青草大香综合精品,日韩美aaa特级毛片,欧美成人精品午夜免费影视

To deal with the problem of large deviation of single sensor in the human motion attitude recognition monitoring, an human motion recognition method based on data fusion between acceleration transducer and gyroscope is proposed. In the method, the attitude angle output by acceleration transducer is modified by the human motion information output by gyroscope, multi-sensor data fusion is achieved by using Kalman filtering algorithm to effectively improve the accuracy of attitude angle measurement. According to the daily movement of human body, Hidden Markov Model based on attitude angle feature is built for human motion recognition. Experiments show that, the proposed method performs better with higher accuracy rate than method using single sensor, which can effectively distinguish between different daily activities.