초록

용어보기논문에서 용어와 풀이말을 자동 추출한 결과로, 시범 서비스 중입니다. AI-Helper

Abstract ▼ AI-Helper

In order to detect a defect in a TFT-LCD image, various research have been continued to analyze a feature between a background pixel and a defective pixel to determine a threshold value. The STD (Standard Deviation) method estimates the intensity distribution of the background based on the mean and ...

In order to detect a defect in a TFT-LCD image, various research have been continued to analyze a feature between a background pixel and a defective pixel to determine a threshold value. The STD (Standard Deviation) method estimates the intensity distribution of the background based on the mean and standard deviation of the image, and determines the threshold value based on the distribution. However, when the size or the intensity of the defect is uncertain, the problem of sensitivity to the change of the weight used for determining the threshold remains. In this paper, we propose a method for robustly measuring the distribution of defective pixels in estimating the intensity distribution of the background. As a method for this, there are the AAD(Average Absolute Deviation) method using the mean of the deviation and the MAD(Median Absolute Deviation) method. In this paper, we propose a method to estimate the background using MAD method. In particular, we propose a method to improve the efficiency of the algorithm by proposing the HMAD(Histogram based Median Absolute Deviation) based on the histogram of the intensity of the TFT-LCD image. As compared with the existing MAD method, the execution time was reduced by 85% and it was faster than the AAD method. Also, we can confirm that the efficiency of the HMAD method increases as the size of the image increases.

Keyword