Developing auto recipe management system for LCD panel auto defect detecting inspection machine

Recently, there is an intensive price competition among mass LCD panel makers. to decrease labor costs, many manufacturing processes such as auto defect detecting are getting automated. However, to maintain its optimal performance, user must keep attention to hardware settings and many recipe parameters. These efforts are supposed to be managed constantly as developing new model. In this paper, we introduce automated defect finding algorithm generally used in periodic pattern image, and suggest auto recipe management algorithms and system that help user to devote less effort for maintenance. Three algorithms are developed - auto pitch calculation, auto threshold control, auto light source intensity control. To verify the performance of auto recipe management algorithm, simulation and machine tests are executed with several color filter and TFT pattern image. Through the tests, we verified the performance of developed algorithm and got successful result.