Passenger abnormal behavior recognition method based on human body posture estimation

Abstract

The invention discloses a passenger abnormal behavior recognition method based on human body posture estimation. The passenger abnormal behavior recognition method based on the human body posture estimation includes the steps of: (1) conducting image acquisition through a monitoring system; (2) dividing a pedestrian passage into multiple areas through the monitoring system; (3) conducting postureestimation of pedestrians according to the acquired image via the monitoring system and acquiring pedestrian postures through the extracting of skeleton information of the pedestrians; (4) via the monitoring system, conducting multi-target tracking of multi-frame images in a corresponding area according to the skeleton information of the pedestrians and conducting behavioral judgment according tomovement trajectories of the pedestrians; (5) the monitoring system issuing an alarm when the pedestrians are judged to be travelling in an abnormal traffic condition based on a normal traffic condition and abnormal traffic condition which are preset according to the different postures and different trajectories of the pedestrians. The passenger abnormal behavior recognition method based on the human body posture estimation has such advantages as accurately issuing an alarm in the case of an abnormal behavior existing in a person being inspected, monitoring and judging quickly, and having widerange of monitoring.

Description

technical field [0001] The invention relates to a passenger behavior recognition method, in particular to a passenger abnormal behavior recognition method based on human body posture estimation. Background technique [0002] In the prior art, in customs, airports, railway stations and other places where there are many people, especially when checking documents, passengers or customers need to queue up in their respective passages for one-on-one inspection. Abnormal behaviors such as crossing, crossing, and tailgating stowaways will affect the inspection order, and need to be identified. Among them, the conventional technology for passenger abnormal behavior recognition method is to set the detection area in the single-lens color image for recognition. This method simply divides the field of view into abnormal areas and normal areas on the single-lens image, and uses background difference If it cannot be judged that the passenger enters the abnormal area, it means that the p...

AI Technical Summary

Problems solved by technology

At present, the background subtraction method to extract the foreground inevitably exists due to problems such as light, lens white balance, etc., and even the luggage carried by passengers will cause interference.

Images