Corrugated board quality detection method and system based on image recognition

By employing multimodal image fusion and convolutional neural networks, a panoramic quality inspection and process optimization for corrugated cardboard was achieved, solving the problems of insufficient real-time performance and data-driven nature of traditional inspection methods, and improving the accuracy of inspection and production efficiency.

CN122391144APending Publication Date: 2026-07-14GUANGZHOU YIWANG PRINTING & PACKAGING CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
GUANGZHOU YIWANG PRINTING & PACKAGING CO LTD
Filing Date
2026-04-17
Publication Date
2026-07-14

AI Technical Summary

Technical Problem

Traditional corrugated cardboard quality inspection relies on manual visual inspection and offline sampling, which cannot achieve real-time quality control. As a result, quality improvement depends on experience and cannot achieve data-driven process optimization and closed-loop control.

Method used

An image recognition-based approach is adopted, which combines multimodal image fusion and convolutional neural networks with geometric calculations to accurately identify surface and internal defects of corrugated cardboard, and constructs a defect-process cause-effect graph for quality assessment and process optimization.

Benefits of technology

It enables panoramic, all-around inspection of corrugated cardboard quality, improving the accuracy and comprehensiveness of defect identification. The objective assessment based on quantifiable data can proactively guide process optimization, reduce scrap rates, and improve production efficiency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122391144A_ABST
    Figure CN122391144A_ABST
Patent Text Reader

Abstract

The application relates to the technical field of image processing, and discloses a corrugated board quality detection method and system based on image recognition. The method comprises the following steps: inputting a registration surface paper color image into a first convolutional neural network to identify surface defects of the corrugated board; inputting a registration transmission light image into a second convolutional neural network to identify internal structural defects of the corrugated board, and calculating global thickness uniformity and local flatness of the corrugated board; calculating a comprehensive quality index of the corrugated board; and inputting the surface defects, the internal structural defects, the global thickness uniformity and the local flatness as observation evidence into a pre-constructed defect-process causal diagram to generate defect process parameters of the corrugated board, combining the comprehensive quality index and the defect process parameters to construct a quality detection report of the corrugated board. The application can improve the reliability of corrugated board quality detection.
Need to check novelty before this filing date? Find Prior Art