Pavement crack detection method and device, electronic equipment and storage medium

Abstract

The embodiment of the invention provides a pavement crack detection method and device, electronic equipment and a medium. The method comprises the steps of collecting a road surface image of a road; preprocessing the pavement image to obtain a plurality of input images with gradient change of resolution; inputting the plurality of input images into a pre-trained pavement crack detection model to obtain a calculation result, wherein the pavement crack detection model is obtained by training based on a sample pavement image and crack mark data of the sample pavement image, the pavement crack detection model comprises a plurality of stages, and multi-time multi-scale fusion is carried out between every two stages; and outputting a detection result according to a calculation result of the pavement crack detection model. According to the embodiment of the invention, by inputting a plurality of input images with gradient change of resolution and using the pavement crack detection model witha multi-scale fusion structure, pavement crack detection under complex conditions is achieved, the noise influence is weakened, and the detection precision is improved.

Description

technical field [0001] The invention relates to the technical field of traffic road image detection, in particular to a road surface crack detection method, device, electronic equipment and storage medium. Background technique [0002] Pavement cracks are one of the main forms of road pavement disease. Early cracks will not affect the use of the pavement, but with the intrusion of rainwater and the action of external forces, it may cause structural damage to the pavement and pose a threat to driving safety. The detection and repair of pavement cracks is an important part of road maintenance. [0003] With the development and popularization of artificial intelligence, related detection methods based on deep learning are gradually used in the task of pavement crack detection. Deep learning does not require manual design of feature extractors, and can automatically perform abstract expressions based on original image features to determine whether a specific pixel is is part o...

AI Technical Summary

Problems solved by technology

[0004] However, the existing methods have the following problems: they only focus on the local characteristics of cracks, which are easily affected by noise; the spatial relationship and connectivity characteristics of cracks are not considered, and the detection results are discontinuous or have isolated points; the detection task is single, and the cracks are not considered. Thickness, inaccurate for crack detection in complex situations

Images