Strip steel complex wave defect detection and early warning method

By using computer vision technology and a three-point angle discrimination method, the problems of missed detection and false detection of complex wavy defects in strip steel in existing technologies have been solved, realizing efficient wavy defect detection and early warning, and ensuring product quality.

CN117816755BActive Publication Date: 2026-07-17BAOSHAN IRON & STEEL CO LTD
View PDF 4 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
BAOSHAN IRON & STEEL CO LTD
Filing Date
2022-09-29
Publication Date
2026-07-17

AI Technical Summary

Technical Problem

Existing flatness testers are unable to effectively detect and identify complex waviness defects in strip steel, especially thin strip steel, which is prone to missed or false detections, affecting product quality.

Method used

A strip steel complex wavy defect detection and early warning system based on computer vision technology is adopted. The system uses a camera to collect data, separates the strip steel from the background through an instance segmentation algorithm, and combines the ray-drawing method and the three-point angle discrimination method to realize the detection of strip steel edges and the identification of wavy defects.

Benefits of technology

It enables the detection of wavy defects across the entire width of the strip, improving the effectiveness and accuracy of the detection, enabling timely detection of wavy defects, ensuring stable product quality, and reducing computational resource consumption.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117816755B_ABST
    Figure CN117816755B_ABST
Patent Text Reader

Abstract

The application discloses a strip steel complex wave defect detection and early warning method, and the steps of the method are as follows: step 1, video data acquisition, a data analysis module extracts real-time video stream of a scene from a data acquisition module and converts the real-time video stream into image data; step 2, image data instance segmentation, a strip steel is separated from an image background through an instance segmentation algorithm, and a coordinate point position set of a strip steel edge in the image is obtained; step 3, tail area point screening, whether the strip steel edge point is in the tail of the strip steel is analyzed and judged, and the points in the tail area are screened out; step 4, wave defect discrimination, whether the included angle θ formed by adjacent three points of the tail area points of the strip steel is between 【A1, A2】 is judged, and when the included angle is greater than two, it is judged that the strip steel has the wave; and step 5, wave defect early warning, if it is judged that the wave defect is generated, an alarm signal is sent to a basic automation computer. The application can find the strip steel with the wave defect, and ensures that the product quality is stable and controlled.
Need to check novelty before this filing date? Find Prior Art