Crack detection method for road surface

Abstract

The invention discloses a crack detection method for a road surface, the method includes: extracting the area of interest of road surface in the image in advance, detecting the image segment endpoint being possible as crack of road surface, adopting multiscale combined with the gradient histogram and support vector machine for classifying, outputting the information of the area, which is judged as the crack of road surface and reminding the user. The video information content is more abundant, the classifying operation is more accurate and the false alarm rate is lower; the method is suitable for larger range with higher automatic processing level, the detection accuracy rate is improved, the detection discriminating efficiency is improved, the method has higher stability, higher anti-noise capability and anti-tilt capability compared with the conventional method; the method can solve the problems that the artificial patrolling work intensity is big at present, the operation is limited by the experience level and the misjudgement and omission rate is high.

Description

technical field [0001] The invention relates to the technical field of automatic video detection, in particular to a method for detecting cracks on highway pavement. Background technique [0002] In recent years, the construction speed and operation scale of my country's high-grade highways have reached the world's leading level. However, with the increase of vehicle speed and the widespread use of heavy-duty vehicles, the excitation frequency of high-grade road subgrade systems has also become wider, and it is susceptible to high-frequency excitation and resonance, causing increased vibration and accelerated aging, which will affect road operation. quality and increase maintenance costs, and will accelerate the structural fatigue damage of pavement asphalt. If the damaged highway pavement is not found and repaired in time, the damage will be accelerated under the action of vehicle rolling, which will affect road comfort and even endanger driving safety. In this situation,...

AI Technical Summary

Problems solved by technology

[0003] However, due to the special problem of information loss caused by the projection of 3D information into 2D images in the process of video imaging, it is still difficult to understand scenes and identify problems based on video images.

At the same time, because the effect of image acquisition in practical applications is limited by the intensity and uniformity of the light and the interference of the complexity of the environment, it will be difficult to distinguish the pavement cracks to be detected from other areas of the pavement (such as figure 1 shown)

Images