Methods, systems and apparatus for defect detection and classification

摘要

Aspects of the present invention are related to systems, methods and apparatus for image-based automatic detection of a defective area in a flat panel display and classification of the defect type and the cause of the detected defect.

主权项

1. A non-transitory computer-readable medium encoded with a computer program code for implementing a method for identifying a defect type associated with a detected defect in a flat panel display, said method comprising:
receiving, in a computing device, a defect mask image associated with an image of a detected defect in a flat panel display; receiving, in said computing device, a plurality of landmark mask images, wherein:
each landmark mask image in said plurality of landmark mask images is associated with one of a plurality of landmarks within a flat panel display; andeach landmark mask image in said plurality of landmark mask images is extracted from a defect-free model image; computing a source-defect connectivity measure between said detected defect and a first source line using a first landmark mask image, from said plurality of landmark mask images, associated with said first source line; computing a gate-defect connectivity measure between said detected defect and a first gate line using a second landmark mask image, from said plurality of landmark mask images, associated with said first gate line; when said source-defect connectivity measure is not equal to zero or said gate-defect connectivity measure is not equal to zero, identifying a defect type associated with said detected defect based on said source-defect connectivity measure and said gate-defect connectivity measure; and when said source-defect connectivity measure is equal to zero and said gate-defect connectivity measure is equal to zero:
computing a first distance measure between said detected defect and said first source line;computing a second distance measure between said detected defect and a second source line using a third land mark mask image, from said plurality of landmark mask images, associated with said second source line;combining said first distance measure and said second distance measure to obtain a source-defect distance measure;computing a gate-defect distance measure between said detected defect and said first gate line; andidentifying said defect type associated with said detected defect based on said source-defect distance measure and said gate-defect distance measure.