Target detection method for vehicles and pedestrians in intelligent traffic monitoring

Abstract

The invention discloses a target detection method for vehicles and pedestrians in intelligent traffic monitoring. The method comprises the following steps of: performing background model initialization on a video frame sequence, independently establishing hybrid Gaussian background component models for a saturation component and a brightness component, and calculating component mean values; differentiating a current frame in the video frame sequence from a background frame, performing binarization processing on a foreground frame, removing shadows and noises and performing morphological filtering; updating weight values, mean values and variances of the components of an obtained hybrid Gaussian background model by using an updating factor; and comparing values of moving target pixels to be matched with each distributed Jeffrey value in the updated hybrid Gaussian background model, and judging whether the moving target pixels are foreground points or not by utilizing the Jeffrey values. By utilizing the method, related parameters of the saturation component and the brightness component are required to be updated, so the overhead of a system is decreased under the condition of not influencing the precision, and the influence of noises and environmental illumination can be avoided; and the method can adapt to the slight disturbance of scenes, and has the characteristic of high robustness.

Description

technical field [0001] The invention belongs to the technical field of digital image processing, and in particular relates to a detection method for vehicles and pedestrians in intelligent traffic monitoring, which detects abnormal moving objects in input traffic videos. Background technique [0002] Intelligent video surveillance is to automatically analyze the image sequence through the computer without human intervention, to realize the operations of moving target detection, tracking and behavior understanding in the dynamic scene, and to judge whether to issue an alarm according to the analysis results. In an intelligent video traffic monitoring system, the moving objects of pedestrians and vehicles in the monitoring scene are the basis of image analysis, image recognition and image understanding. The results of moving object detection can be used for subsequent object tracking and classification, and the detection effect is direct. It will affect the follow-up work, so ...

AI Technical Summary

Problems solved by technology

[0004] Aiming at the problems of sensitive illumination, ghost phenomenon and unremoved shadows in the detection of moving objects in the current intelligent video processing technology, the present invention proposes an improved method for detecting moving objects, which overcomes the problem of applying the mixed Gaussian background model to traffic intelligent video. The deficiencies in monitoring, to detect the targets of vehicles and pedestrians in intelligent traffic monitoring, effectively improve the sensitivity and accuracy of moving target detection

Images