Human Posture Recognition Based on Isolation Forest Algorithm
Hao Ge, Lizuo Jin · Advances in transdisciplinary engineering · 2023
With the rapid development of semiconductor technology and mobile communication networks, inertial sensors have higher performance, and correspondingly, the demand for human behavior recognition is increasing. At this stage, the mainstream research methods mainly analyze images in video streams. Although they can achieve good recognition results, due to the high requirements for camera location, environmental conditions, and privacy considerations, it is difficult to popularize behavior recognition methods for videos. Therefore, this article conducts research on human behavior recognition based on wearable sensor data, The innovative introduction of the isolation forest algorithm for prior processing of data can quickly detect anomalies in the behavior process, and also reduce data anomalies caused by the impact of sensor performance. Finally, the Convolutional Neural Networks is used for classification to obtain the final recognition results. The system has been tested on the MHEALTH datasets, and the recognition rate has reached 99.01%, achieving good classification results.