Fog penetration identification method for an intelligent traffic monitoring image

Abstract

The invention discloses a fog penetration identification method for an intelligent traffic monitoring image. The method mainly comprises the following steps: A, acquiring a traffic monitoring image, constructing a video image signal wavelet coefficient, and denoising the image according to the setting of a coefficient threshold; B, constructing a foggy day image degradation model; preliminarily estimating the transmissivity of the image according to a dark channel prior idea, and obtaining a clearer defogged image by adopting a rapid guide filtering method; c, constructing a novel filter basedon phase type distribution and characteristics of a related random model, constructing a consistent approximation method according to the approximation of the filter, and enhancing the response of the filter to image signals; and D, selecting an edge detection optimal criterion to carry out edge detection and segmentation of the image according to consistent approximation of the novel filter, andfinally identifying the image to complete image identification of traffic monitoring. The method has good real-time performance and implantability, the monitoring image can be processed according toimage denoising processing, dark channel defogging processing and edge detection methods, and the fog penetration identification task of the traffic monitoring image is effectively completed.

Description

technical field [0001] The invention relates to an image processing method of an intelligent traffic monitoring system, which belongs to the field of computer and pattern recognition. Background technique [0002] The intelligent traffic monitoring system is to transmit the on-site images in the monitoring area back to the command center through the monitoring system, so that the traffic management personnel can directly grasp the traffic conditions such as vehicle queuing, congestion, and signal lights. In the foggy weather, the monitoring facilities, means and functions of the existing monitoring system are relatively backward. In some hit-and-run traffic accidents, it is impossible to accurately identify the color, number, type and other information of the vehicle through the images in the monitoring video. [0003] In view of the above problems, the present invention adopts the methods of image denoising processing, defogging processing and edge detection to realize the ...

AI Technical Summary

Problems solved by technology

In the foggy weather, the monitoring facilities, means and functions of the existing monitoring system are relatively backward. In some hit-and-run traffic accidents, it is impossible to accurately identify the color, number, type and other information of the vehicle through the images in the monitoring video.

Images