Behavior recognition method, device and computer-readable storage medium

Abstract

The present invention relates to the technical field of image recognition, and specifically provides a behavior recognition method, device and computer-readable storage medium, aiming to solve the technical problem of how to accurately recognize human behavior associated with small targets such as cigarettes / mobile phones. For this purpose, according to the method of the embodiment of the present invention, it is possible to extract the partial image of the human body part related to the behavior to be recognized in the human body image to be recognized, and then perform the behavior to be recognized according to the complete human body image to be recognized and the extracted partial image at the same time. Behavior recognition, that is, behavior recognition based on the global image features and local image features of the human body image to be recognized, which reduces the probability of false detection due to the small size of the cigarette / mobile phone associated with the behavior to be recognized, thereby improving the The recognition accuracy of the behavior to be recognized.

Description

technical field [0001] The present invention relates to the technical field of image recognition, in particular to a behavior recognition method, device and computer-readable storage medium. Background technique [0002] In application scenarios with high safety requirements such as oil fields, gas stations, and driving vehicles, it is necessary to effectively identify whether there are human behaviors that endanger facility / personal safety, such as smoking and calling mobile phones, and identify the existence of the above human behaviors After that, information alarms can be issued in a timely manner. At present, conventional monitoring methods for smoking and calling mobile phones mainly include monitoring methods using smoke alarms / signal jammers and monitoring methods based on image recognition. [0003] Specifically, the method of using smoke alarms / signal jammers to monitor mainly includes using smoke alarms to monitor the concentration of smoke, selectively performin...

AI Technical Summary

Problems solved by technology

However, in order to improve the monitoring accuracy of smoking and calling mobile phones, a large number of devices such as smoke alarms and signal blockers need to be deployed, which will significantly increase the monitoring cost

The monitoring method based on image recognition mainly includes the use of conventional classification algorithms or detection algorithms in the field of machine learning technology to identify the behavior of smoking and calling mobile phones on images. The false detection rate of the algorithm increased, which eventually led to a decrease in the recognition accuracy of smoking and calling mobile phones

Images