A pedestrian intelligent identification method and system thereof

Abstract

The invention provides a pedestrian intelligent identification method and system thereof, which includes a video data collection step, a pedestrian parameter feature extraction step and a pedestrian re-identification step. Wherein the system for implementing the pedestrian intelligent identification method involved in the present invention includes 2 or more cameras, front-end processors and servers. Wherein the camera is used to implement the video data collection step, and the front-end processor is used to manage the cameras, and implement the pedestrian parameter feature extraction step and the pedestrian sample video extraction step according to the video data acquired by the camera. The server is used to save the pedestrian data information involved in the surveillance video obtained by the front-end processor, and to implement the corresponding pedestrian re-identification step.

Description

technical field [0001] The present invention relates to a pedestrian recognition method and system thereof, in particular, a pedestrian intelligent recognition method and system thereof based on pedestrian movement trajectories. Background technique [0002] With the continuous development of society, the general public's demand for safety is also getting higher and higher. In response to this demand, relevant departments are also constantly strengthening their strength to maintain social security. [0003] At the same time, with the continuous development and progress of video surveillance technology, relevant departments have also introduced it into the field of public security. Due to its video recording, playback and other characteristics, it plays an important role in social public security. getting bigger. [0004] With the continuous popularization of video surveillance systems, the areas where they are deployed are becoming more and more extensive, which obviously ...

AI Technical Summary

Problems solved by technology

Therefore, the movement trajectory of a specific pedestrian in the surveillance video is incomplete, and its behavior characteristics cannot be well described, so that the subsequent intelligent analysis of pedestrian trajectory cannot be carried out effectively.

However, if new surveillance video is added, the efficiency of video data analysis is obviously very low because currently only relying on manpower to connect and analyze relevant video data

Images