Pedestrian head lowering abnormal behavior detection method based on human body articulation points

Abstract

The invention discloses a pedestrian head lowering abnormal behavior detection method based on human body articulation points, which comprises the following steps of: 1, acquiring each frame of imagein a video stream, expanding the video stream into an image sequence, detecting each pedestrian in a video frame by adopting a human body detection algorithm for each frame in the sequence, and cutting a human body region frame; 2, independently estimating the posture of each pedestrian in the generated human body region frame by adopting a human body posture algorithm; 3, mapping the estimated human body posture back to the original image, and finely adjusting a human body region frame until accurate coordinate information of human body articulation points is generated; 4, repeating the step1, the step 2 and the step 3 until the position of the human body area frame does not change any more; 5, during training, adopting an abnormal behavior fitting algorithm to fit the accurate human body joint point information with the normal behavior into the abnormal behavior, generating a negative sample, taking the normal behavior before fitting as a positive sample, forming a training set by the positive sample and the negative sample to train a behavior classifier, and generating an optimal model; and 6, during testing, directly sending the generated coordinate information of the accuratearticulation points into the trained behavior classifier, and detecting and outputting a result. According to the invention, the problem of low accuracy of abnormal behavior detection of a pluralityof pedestrians in a complex environment is solved, and an effect of high-accuracy multi-person real-time detection in a complex scene is realized.

Description

technical field [0001] The invention relates to the technical field of traffic safety management, in particular to a method for detecting the abnormal behavior of pedestrians bowing their heads based on human body joints. Background technique [0002] The intelligent development of mobile phones has made people invest more and more energy, but when pedestrians cross a crosswalk or wait for traffic lights, if they look down at their mobile phones, it is easy to distract their attention and reduce their reaction speed to external dangers. According to the World Health Organization, more than 270,000 pedestrians die in road traffic accidents every year around the world, and the use of mobile phones on the way is the main reason for traffic accidents. Countries around the world have taken various measures to rectify bowing people, but the response has been mediocre. Therefore, it is imminent to automatically detect and identify the abnormal behavior of lowering the head of pede...

Images