Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
MACHINE VISION-BASED QUALITY INSPECTION METHOD AND SYSTEM UTILIZING DEEP LEARNING IN MANUFACTURING PROCESS
Document Type and Number:
WIPO Patent Application WO/2019/107614
Kind Code:
A1
Abstract:
A machine vision-based quality inspection method and system utilizing deep learning in a manufacturing process are provided. A quality inspection method according to an embodiment of the present invention: generates a learning product image; enables a classifier to be learned for classifying a good product and a defective product through the generated learning product image; and determines a product as a good product or a defective product by using the learned classifier. Therefore, it is possible to find the feature value of data of a product to be classified on the basis of learning, such that it is possible to carry out the machine vision-based inspection even on an inspection area that relies on a manual inspection due to the difficult of formalizing defects.

Inventors:
LEE SEUNG WOO (KR)
KWON YOUNG MIN (KR)
Application Number:
PCT/KR2017/013922
Publication Date:
June 06, 2019
Filing Date:
November 30, 2017
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
KOREA ELECTRONICS TECHNOLOGY (KR)
International Classes:
G06T7/00; G01N21/88; G06K9/32; G06K9/62; G06T7/11
Foreign References:
KR100786823B12007-12-20
KR101688458B12016-12-23
JP2005156334A2005-06-16
KR20100088410A2010-08-09
KR20160105082A2016-09-06
Attorney, Agent or Firm:
NAM, Choong Woo (KR)
Download PDF: