A machine vision-based surface defect recognition system and method in straw production

CN120707558BActive Publication Date: 2025-11-07YIWU SHUANGTONG DAILY NECESSITIES CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202511127331.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2025-08-13
Publication Date
2025-11-07
Estimated Expiration
2045-08-13

AI Technical Summary

Technical Problem

In straw production, existing machine vision inspection technology struggles to effectively segment tiny or low-contrast surface defects in complex backgrounds, and deep learning models have high requirements for diverse training data and computation, resulting in insufficient detection efficiency and accuracy.

Method used

By extracting the center line of the eyedropper for pose alignment and normalizing illumination artifacts, a normalized image is generated. Combined with a preset standard template, a defect-enhanced residual heatmap is generated. Finally, a segmentation neural network with a coordinate attention mechanism is used for defect segmentation and classification.

Benefits of technology

It enables precise and automated identification of defects on the surface of straws, improves the segmentation accuracy and identification accuracy of small, irregular and low-contrast defects, and meets the real-time inspection needs of high-speed production lines.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120707558B_ABST
    Figure CN120707558B_ABST
Patent Text Reader

Abstract

The application relates to the technical field of image analysis, in particular to a straw production surface defect recognition system and method based on machine vision, which comprises the following steps: an image acquisition module acquires an original RGB image of a straw; a pretreatment and feature enhancement unit processes the original image, including extracting a straw center line, performing posture alignment and illumination artifact normalization based on the center line to generate a normalized image, and fusing a defect enhancement residual heat map by calculating a symmetry residual component of the normalized image and a template difference component based on a preset standard straw template; a defect segmentation and classification unit generates an image defect mask by using a segmentation neural network containing a coordinate attention mechanism, and performs feature extraction and classification on the mask area to determine the defect type; finally, a real-time output unit outputs the position and confidence information of the defect. The application aims to realize accurate and automatic recognition of straw surface defects through multi-stage image analysis and processing.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Straw detection system and method

    CN111222445A

  • Target tracking method and device based on anchor-point-free twin network

    CN113112523A