Title:
ARTIFICIAL INTELLIGENCE-BASED METHOD AND SYSTEM FOR DETECTION AND CLASSIFICATION OF SEWING DEFECT
Document Type and Number:
WIPO Patent Application WO/2024/071670
Kind Code:
A1
Abstract:
Disclosed are an artificial intelligence-based method and system for detection and classification of a sewing defect. In a computer system, a large amount of prepared training images for each type of sewing defect are learned by using an artificial neural network-based machine learning algorithm or a deep learning algorithm so as to obtain a sewing defect detection and defect type classification model. In the computer system, an artificial neural network-based computer program having the sewing defect detection and defect type classification model applied thereto is executed, and real-time captured images of a workpiece being sewn provided from a camera are input into the sewing defect detection and defect type classification model, so that sewing defects that occur during a sewing operation are detected in real time and the type of sewing defect is classified. If a defect is detected, information about the detected defect, including measures appropriate to the type of defect, is displayed to an operator.
More Like This:
Inventors:
JUNG WOO-KYUN (KR)
LEE JAEWON (KR)
LEE JAEWON (KR)
Application Number:
PCT/KR2023/011999
Publication Date:
April 04, 2024
Filing Date:
August 11, 2023
Export Citation:
Assignee:
MFAFA CO LTD (KR)
International Classes:
D06H3/08; D05B19/00; G01N21/88; G06N3/08; G06T7/00; H04N7/18
Foreign References:
| CN111028204A | 2020-04-17 | |||
| KR102322546B1 | 2021-11-09 | |||
| CN110175988A | 2019-08-27 | |||
| US20220044389A1 | 2022-02-10 | |||
| CN114066820A | 2022-02-18 |
Attorney, Agent or Firm:
PARK, Young-Woo (KR)
Download PDF: