학술논문
Automated Bright Pixel Detection System on LCD Displays
Document Type
Conference
Author
Silva, Mateus O.; Ferreira, David A. O.; Ouchi, Kethilen Y.; Torres, Gustavo M.; Urtiga Mattos, Edma V.C.; Pereira, Antonio M. C.; Costa, Luciana R.; Cavalcante, Victor L. G.; Cruz, Caio F.S.; Silva, Agemilson P.; Belem, Ruan J.S.; Jesus, Anderson S.; Junior, Wilson C.C.; Paula, Ricardo G.; Bezerra, Thiago B.; Junior, Waldir S. S.; Carvalho, Celso B.
Source
2021 IEEE International Conference on Consumer Electronics (ICCE) Consumer Electronics (ICCE), 2021 IEEE International Conference on. :1-4 Jan, 2021
Subject
Language
ISSN
2158-4001
Abstract
Bright pixel defect occurs in LCDs when a transistor of a subpixel remains bright. Bright pixels reduce the quality of manufactured devices, negatively impacting companies looking to meet an increasingly demanding market. In this wokr, conducted by three partners, called UFAM/CETELI, Envision (TPV Group) and ICTS, we propose an automated bright pixel defect detection system on LCD displays, which works in conjunction with two proposed methods, the quartile-based method and the standard-based method to identify bright pixels in LCDs. The proposals obtained as best result 100% of accuracy and between 0.397s and 0.193s of average processing time per image. The results of high accuracy and reduced processing time demonstrate the contribution of our proposal, being possible to use it in the detection of bright pixels in real production lines that use LCDs on TVs or monitors.