A Human Activity Recognition Method For Wearable Device

With the development of science and technology and the continuous improvement of manufacturing technology, smart phones as the representative of intelligent devices continue to innovate and almost affect all aspects of people's lives. At the same time, people pay more attention to their daily exercise state. The existing research on the human body attitude and activity is based on the research of wearable devices. Therefore, the research of human body posture recognition based on smart wearable devices will be an important direction for the future development.In the paper, a smart phone within the block tri-axial acceleration sensor is used to simulate wearable device simulation as data acquisition facility. The daily motion of the human body and motion capture 3D acceleration data and related method is proposed to four daily basic posture(walking, running, upstairs and downstairs) recognition, and further on the conversion between various activity analysis, aiming at special fall situation recognition. The method proposed in this paper has a good recognition effect on the recognition of the basic activity and fall condition.The main contents and contributions of this paper are as follows:1. Based on the analysis of the human stepped stress, the paper analyse each stride interval and put forward a method based on single stride interval feature selection which makes the study more detailed and more accurate in location.2. Puts forward a classification and recognition method based on single stride interval feature vector which can be more effective to identify the four basic stride stance(walk, run, upstairs and downstairs).3. Puts forward an effctive recognition method of fall. The method combines with identified walking pace,and compares the conversion interval Accurate which has differdent continuous stance before and after accurate positioning...