Method for splicing and homogenizing road face images collected by double-linear-array cameras under linear laser illumination condition

Abstract

The invention discloses a method for splicing and homogenizing road face images collected by double-linear-array cameras under the linear laser illumination condition. Two linear-scanning cameras are mounted on the top of the tail of a vehicle, a linear laser is used for illumination, a mileage sensor is used for triggering the cameras to shoot road face images, images obtained by the two cameras are spliced to form one image, lane line and running mark detection is carried out on the image, image gamma correction, namely homogenizing, is carried out on non-lane-line and non-running-mark areas in the image, and image intensification is carried out finally. The real time performance of the method can meet the requirement for real-time collection and real-time homogenizing, interference on image homogenizing of lane lines and running marks can be omitted in the homogenizing effect, and the robustness can be adaptive to images of various road faces like expressways, provincial level roads and municipal roads.

Description

technical field [0001] The invention relates to the field of image acquisition under line laser lighting conditions, and is a double-camera acquisition image splicing method and an image homogenization method under line laser lighting conditions, especially a road pavement image homogenization method, which is especially widely used in highways and provincial Various road surface disease detection fields such as highways and municipal roads. Background technique [0002] With the increase of highway mileage in our country, the supporting highway maintenance management has also become a complex system engineering, especially the maintenance task of highway, which has become one of the key factors to ensure its normal operation; If it is discovered and repaired, the maintenance cost of the road surface will be greatly reduced, and it will also provide a reliable guarantee for safe driving; how to conduct a comprehensive inspection and positioning of the entire road surface saf...

AI Technical Summary

Problems solved by technology

[0003] The commonly used detection method in the past is to use artificial visual observation, which is inefficient, affects traffic, and is highly dangerous, and there are large subjective differences in the detection results, and the accuracy is not high. At present, with the expansion of the road network, the manual method can no longer meet the requirements. ; Therefore, the automatic detection method of road surface defects based on digital images has been widely studied and used to overcome the above defects. A series of intelligent and efficient road detection equipment makes it possible to obtain road information on a large scale, quickly and accurately. The detection equipment generally consists of Composed of proprietary vehicles and on-board sensors, the software generally consists of a road surface image acquisition module and a later-stage disease identification and processing module, which can complete the detection of the road without affecting normal traffic, and grasp the road conditions in a timely and accurate manner. Provides an important basis for formulating conservation measures

[0004] Road surface image acquisition is the premise of later disease identification processing, and the quality of the collected images also directly affects the accuracy and effect of identification; currently, road surface image acquisition generally adopts dual line array cameras with line laser lighting, because the standard lane width is 3.75m, requiring one The detection can cover the entire lane, and the dual-camera method can reduce the deformation of the image and improve the detection accuracy of the disease. Generally, it is difficult for the scanning lines of two line-scan cameras to overlap accurately, so the direct splicing of the scanned images will cause the left and right images to have a sense of misalignment. And generally there are some overlapping images in the middle, how to quickly and automatically complete the automatic registration of the two images and accurately stitch them together is one of the problems that need to be solved; in addition, due to the inhomogeneity of the current line laser (compared with natural light) , so that the image captured by the camera has a large number of light and dark stripes, resulting in poor visual effect of the image, which will also seriously affect the later disease identification. How to effectively eliminate such light and dark stripes is another problem that needs to be solved

Images