Dynamic Detection Method of Pavement Cracks Based on Video Stream

Abstract

The invention discloses a pavement crack dynamic detection method based on a video stream. The pavement crack dynamic detection method comprises the following steps that step 1, the crack area of theroad pavement is enclosed in a rectangle frame, the video of the road pavement including the crack area is acquired by using a vehicle-mounted motion camera, any one frame of video is extracted to actas a video image and grayscale processing is performed on the video image so as to obtain a grayscale image; step 2, binarization processing is performed on the grayscale image so as to obtain a binary image; step 4, area growing algorithm processing is performed on the crack area of the binary image so as to obtain the processed crack area; step 5, the processed crack area is corrected so as toobtain the corrected crack area; and step 6, the area of the crack in the corrected crack area is calculated. The disadvantages that the manual detection method has high labor intensity, low safety, traffic interference, low work efficiency and low detection accuracy can be overcome.

Description

technical field [0001] The invention belongs to the field of road engineering, in particular to a method for dynamically detecting cracks on a road surface based on video streams. Background technique [0002] During the long-term operation, the expressways built in the early days were continuously damaged by factors such as the natural environment and human travel load, and various road surface damages inevitably occurred. Therefore, it is necessary to timely and accurately count the damage of the road surface, and formulate corresponding reasonable and scientific methods to maintain the road, so as to avoid the continued expansion of the disease and endanger people's travel safety. Therefore, it is necessary to make accurate statistics on the damage factors such as the type and area of ​​pavement damage, analyze and evaluate the pavement damage status, so as to provide a large amount of accurate and timely data basis for the traffic management department to carry out road ...

AI Technical Summary

Problems solved by technology

[0005] In order to solve the defects of manual visual inspection of road conditions and the cost of fully automatic inspection vehicles, and to increase the frequency of damage detection on roads of all levels, a dynamic detection method for road cracks based on video streams is proposed.

Images