A highway surface crack rapid detection method and system based on high-resolution image recognition

By using high-resolution image recognition technology, image block offset correction and manual repair area screening, the problems of false alarms and missed alarms in traditional detection methods have been solved, and accurate detection and quantitative assessment of highway cracks have been achieved.

CN121305350BActive Publication Date: 2026-06-09GUANGDONG TIANYUAN TECHNOLOGY CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
GUANGDONG TIANYUAN TECHNOLOGY CO LTD
Filing Date
2025-10-16
Publication Date
2026-06-09

AI Technical Summary

Technical Problem

Traditional methods for detecting cracks on highways are susceptible to interference from complex road surface environments, leading to false alarms and missed alarms. They also lack the ability to deeply analyze crack morphology and texture, making it difficult to accurately measure key geometric parameters of cracks, thus affecting the reliability and accuracy of detection results.

Method used

A high-resolution image recognition method is adopted, which generates a corrected composite image by image block offset correction, manual repair area screening and recognition model, extracts and distinguishes real crack information, determines the final crack type and extracts the corresponding information.

Benefits of technology

It effectively eliminates image misalignment and deformation, reduces false alarm rate, improves the accuracy and reliability of detection results, enables precise classification of crack types and quantitative assessment of damage degree, and provides detailed decision-making basis for road maintenance.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121305350B_ABST
    Figure CN121305350B_ABST
Patent Text Reader

Abstract

The application belongs to the technical field of highway crack detection, and particularly relates to a highway surface crack rapid detection method and system based on high-resolution image recognition. The method divides a target image into image blocks with overlapping areas, calculates alignment deviation values according to the overlapping areas and performs offset correction to generate a corrected composite image. Initial crack information is extracted from the corrected composite image, and repaired cracks are screened out according to artificial repair areas to generate an image to be evaluated. The image to be evaluated is input into a recognition model to determine the final crack type and extract corresponding final crack information. The application eliminates image misplacement through offset correction, improves data geometric accuracy, significantly reduces the false positive rate by screening out repaired crack information, and simultaneously realizes accurate classification and quantitative evaluation of cracks, thereby providing a scientific decision basis for road maintenance.
Need to check novelty before this filing date? Find Prior Art