SYSTEM FOR DETERMINING DEFECT IN DISPLAY PANEL ON BASIS OF MACHINE LEARNING MODEL

A system for determining a defect in a display panel on the basis of a machine learning model, according to the present invention, comprises: an optical inspector which obtains image information about whether a display panel is defective, by using a lighting inspection method, and classifies the obt...

Full description

Saved in:
Bibliographic Details
Main Authors YUN, Byoung Kook, KIM, Mi Jin, PARK, Nam Jae, KIM, Tae Sung, LEE, Ji Seob, JANG, Sae Il
Format Patent
LanguageEnglish
French
Korean
Published 28.07.2022
Subjects
Online AccessGet full text

Cover

Loading…
More Information
Summary:A system for determining a defect in a display panel on the basis of a machine learning model, according to the present invention, comprises: an optical inspector which obtains image information about whether a display panel is defective, by using a lighting inspection method, and classifies the obtained image information into one of a plurality of defect patterns; and a defect determiner which determines whether the display panel is defective, by applying the obtained image information to a machine learning model, corresponding to the classified defect pattern, among machine learning models respectively corresponding to the plurality of defect patterns. Un système de détermination d'un défaut dans un panneau d'affichage en fonction d'un modèle d'apprentissage automatique, selon la présente invention, comprend : une unité d'inspection optique qui obtient des informations d'image concernant la défectuosité ou non d'un panneau d'affichage, à l'aide d'un procédé d'inspection d'éclairage, et classifie les informations d'image obtenues dans un motif d'une pluralité de motifs de défaut ; et un dispositif de détermination de défaut qui détermine si le panneau d'affichage est défectueux ou non, par l'application des informations d'image obtenues à un modèle d'apprentissage automatique, correspondant au motif de défaut classifié, parmi des modèles d'apprentissage automatique correspondant respectivement à la pluralité de motifs de défaut. 본 발명에 따른 머신러닝 모델에 기반한 디스플레이 패널의 결함 판정 시스템은 점등 검사 방식을 이용하여 디스플레이 패널의 결함 여부에 대한 이미지 정보를 획득하되, 상기 획득한 이미지 정보에 대해 복수의 결함 패턴들 중 어느 하나의 결함 패턴으로 분류하는 광학 검사기; 및 상기 복수의 결함 패턴들에 대응하는 각각의 머신러닝 모델들 중 상기 분류된 결함 패턴에 대응하는 머신러닝 모델에 상기 획득한 이미지 정보를 인가하여 상기 디스플레이 패널에 대한 결함 여부를 결정하는 결함 판정기를 포함하는 것을 특징으로 한다.
Bibliography:Application Number: WO2021KR01071