Transverse bright line defect detection method, device, equipment and storage medium
CN122312632BActive Publication Date: 2026-08-28CREATOR CHINA TCH CO +1
Patent Information
- Application Number
- CN202610771716.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2026-06-01
- Publication Date
- 2026-08-28
- Estimated Expiration
- 2046-06-01
AI Technical Summary
Technical Problem
[0004]本申请的主要目的在于提供一种横向亮线缺陷检测方法、装置、设备及存储介质,旨在解决如何准确对横向亮线缺陷进行检测的技术问题
Benefits of technology
本实施例提出的一种横向亮线缺陷检测方法,获取原始图像信息;基于所述原始图像信息中的检测通道信息对预设裁剪区域图像进行处理,得到图像处理信息以及对应的背景信息;基于所述图像处理信息和所述背景信息确定差分信息;基于所述差分信息对各行的能量参数进行解算,确定能量参数信息;基于所述能量参数信息将对应的横向亮线缺陷指标进行迭代检测,直至满足预设终止条件,得到横向亮线缺陷指标检测结果。本申请通过获取原始图像信息,无需额外采集均匀靶标即可直接基于业务图像开展检测,降低设备质检与运维成本,根据原始图像中的检测通道信息对预设裁剪区域图像进行局部对比度增强与横向去纹理处理,得到图像处理信息以及对应的背景信息,有效抑制文字、底纹等复杂纹理干扰,突出横向亮线信号,在此基础上,利用图像处理信息与背景信息做差分,保留正向窄带响应,从而准确分离亮线缺陷与正常背景,避免误检,对差分信息逐行解算能量参数,获得各行能量分布,将二维图像缺陷转化为一维能量序列,便于快速定位异常行,从而根据能量参数信息对横向亮线缺陷指标进行多候选行循环迭代检测,直至满足预设终止条件,输出缺陷位置与置信度指标,有效区分真实亮线与伪异常,在低算力条件下实现高准确率检测,并支持硬件闭环调整或告警。
✦ Generated by Eureka AI based on patent content.
Smart Images

Figure CN122312632B_ABST
Abstract
The application discloses a transverse bright line defect detection method and device, equipment and a storage medium, and relates to the technical field of image processing. The method comprises the following steps: acquiring original image information; processing a preset cropping region image based on detection channel information in the original image information to obtain image processing information and corresponding background information; determining difference information based on the image processing information and the background information; calculating the energy parameters of each row based on the difference information to determine energy parameter information; and iteratively detecting corresponding transverse bright line defect indicators based on the energy parameter information until a preset termination condition is met to obtain a transverse bright line defect indicator detection result. The preset cropping region image is processed to iteratively detect the transverse bright line defect indicators until the termination condition is met, so that the real bright line and the abnormality can be effectively distinguished without additional targets, and the defects can be accurately detected under low computing power conditions.
Need to check novelty before this filing date? Find Prior Art
Citation Information
Patent Citations
Crop yield prediction method and system based on low-altitude remote sensing information from unmanned aerial vehicle
AU2020101709A4
Defect segmentation positioning method and system for inorganic mineral casting image
CN121073980A