Defect detection of aerial images without reference image

Currently, many different defect detection algorithms have developed. Most finds defect using optical method, and some using image processing uses a reference image. There are several problems using the reference image. The scale, illumination or orientation difference between the reference and the target image make it hard or even impossible to find defect by comparison. Also, there might not be a reference image existed from first place. In this paper, we present a method that automatically find and analyze the defects from an aerial image obtained from a commercial mask inspection device, Aerial Image Measurement System (AIMS). By using this, we developed the process create a reference image from the test image with defect, compare the test image with reference image, and find the exact defect position. Our system finds the defect fully automatically in a short time with high accuracy.