The Abnormal Behavior Recognition Based on the Smart Mobile Sensors

The widespread popularity of smart phones with built-in sensor technology makes our life more convenient. In this paper, a new method based on smart phone is proposed to recognize the behavior of users. Firstly, the intelligent mobile phone is used to realize the behavior data collection, then the High Pass Filtering Algorithm and the Mean Filter Algorithm are combined to pre-process of the collected data, finally, we use the method of Multi Strategy Feature fusion to realize data classification and recognition, the last realize the discovery and recognition of mobile phone user's behavior. Experiments through the 6 kinds of behavior, including walking, running, jumping up and down stairs, tumbling, fall, jumping behavior of 30 different performers to complete 1440 sets of 8 sets of experimental data independently, the recognition accuracy is above 90% by using the method mentioned above, the experimental results show that the proposed method is effective.