Image quality evaluation method of road traffic closed-circuit television based on machine vision

Abstract

The invention discloses an image quality evaluation method of a road traffic closed-circuit television (Closed-Circuit Television, CCTV) based on machine vision, specifically comprising: displaying preset standard images through a portable imaging device; after carrying out a complete set of steps of the CCTV system, outputting real images at a display terminal, wherein the complete set of steps comprises collection, coding, transmission, decoding and so on; photographing over against the display terminal of the CCTV system by a high definition industrial camera to obtain test images; respectively extracting vision features from the test images and the standard images by an information processing unit, comparing and analyzing the vision features, calculating the values of a series of image quality evaluation indexes; resolving the overall object evaluation results of the image quality through an image quality multi-index fusing model by the information processing unit. According to the invention, comprehensive evaluation is carried out to the complete set of CCTV system at the same time without influencing the existing device operation; the overall object evaluation results of the image quality in accord with subject evaluation can be obtained; the method and the system are simple in principle, low in cost and convenient in operation.

Description

technical field [0001] The invention relates to the field of video monitoring in road traffic safety, in particular to a method for evaluating the image quality of road traffic closed-circuit television based on machine vision. Background technique [0002] In road traffic, closed-circuit television (Closed-CircuitTelevision, CCTV) is a beneficial supplement to other information collection equipment such as vehicle detectors. Intuitively obtain the on-site images of traffic jams and upload them to the monitoring room, enabling operation managers to quickly and accurately identify the type and severity of accidents, thus providing a reliable basis for the formulation of traffic emergency and control strategies. [0003] Video image quality is a key technical index to measure the quality of road traffic CCTV monitoring system. In practical applications, there are many factors that affect the image quality of the system. Low camera resolution, inappropriate lens focal length, ...

AI Technical Summary

Problems solved by technology

Among them, the subjective evaluation method is the most reliable because people are the final consumers of images, but there are also obvious shortcomings: they are easily affected by the background knowledge, observation motivation and emotional state of the evaluators, and cannot achieve repeatable and accurate measurements; In order to make the subjective evaluation as accurate and objective as possible, a large number of people are needed as observers, the labor cost is high and the operability is poor

On the other hand, the existing objective evaluation methods can only detect abnormalities in a single link such as video signal acquisition, transmission, storage, encoding and decoding in the CCTV system, or can only use a single evaluation index to Image damage analysis and evaluation, that is, none of the existing related technologies involves comprehensive analysis of the entire CCTV system and an objective image quality evaluation method that considers multiple damage evaluation indicators at the same time

Therefore, the evaluation results cannot be expressed concisely and intuitively like the subjective evaluation method, and it is easy to cause the evaluation results to be inconsistent with the subjective feelings of the image observers

Images