Efficient pedestrian falling detection method based on image enhancement algorithm and human body key point positioning

Abstract

The invention discloses an efficient pedestrian falling detection method based on an image enhancement algorithm and human body key point positioning. The efficient pedestrian falling detection methodcomprises the following steps: (1) extracting pedestrian behavior video clips; (2) processing the image in the video; (3) detecting a person in the video segment; (4) capturing a picture of the tracking frame of the same tracked target person as a human body key point and obtaining position coordinates of 18 key points of the human body; (5) analyzing whether the target person falls down or not according to the positions of the key points in the step (4); (6) judging that the user falls down, storing a video segment, and sending alarm information; and (7) hearing the alarm prompt and taking corresponding measures by a user. According to the method, an image enhancement algorithm and human body key point positioning are combined, existing manpower, material resources and financial cost canbe greatly saved, rapid response is made for timely obtaining pedestrian falling information, and the method has high application requirements and market prospects in the field of social public safety.

Description

technical field [0001] The invention belongs to the technical field of computer vision analysis, and in particular relates to an efficient pedestrian fall detection method based on an image enhancement algorithm and human body key point positioning. Background technique [0002] With the rapid development of the times, my country has gradually entered the aging stage. According to research and surveys, falls are the primary cause of injury and death for 65-year-olds. The occurrence of falls in the elderly is not an accident, as most people think, but a potential danger. Once the elderly fall, if they can be detected in time and rescued as soon as possible, it is possible to effectively reduce the risk of death and long-term hospitalization. Therefore, it is particularly important to detect, prevent and control falls in the elderly. [0003] At present, there are mainly two methods for detecting falls, one is a fall detection method based on wearable sensors, and the other ...

AI Technical Summary

Problems solved by technology

However, this method also has certain deficiencies. For example, the monitoring camera will be affected by different degrees of light and wind, and the video image will be under-exposed or over-exposed and shaken and blurred. In addition, this method usually requires simultaneous Through the analysis of the two models of target detection and feature point positioning, there will be problems of large computer resource consumption and high calculation costs.

Images