Calibration of CMOS Sensor

碩士 === 國立中央大學 === 光電科學與工程學系 === 107 === This paper is mainly to establish a process to correct the uniormity on the CMOS Sensor and improve the quality of the image. A light source is placed in the integrating sphere, and the CMOS sensor to be corrected is incident. The gain value and the offset of...

Full description

Bibliographic Details
Main Authors: Yen-Lin Lee, 李彥霖
Other Authors: Chao-Wen Liang
Format: Others
Language:zh-TW
Published: 2019
Online Access:http://ndltd.ncl.edu.tw/handle/p4pgj6
Description
Summary:碩士 === 國立中央大學 === 光電科學與工程學系 === 107 === This paper is mainly to establish a process to correct the uniormity on the CMOS Sensor and improve the quality of the image. A light source is placed in the integrating sphere, and the CMOS sensor to be corrected is incident. The gain value and the offset of the pixel are obtained by calculating the relationship of the gray scale value of each pixel on the CMOS sensor by controlling the intensity of the light source correcting the CMOS Sensor. In the initial calibration, we corrected the original data at a single shutter time and found that uniformity of the corrected data was significantly improved. However, during the experiment, it was found that the relationship of the grayscale value of each pixel on the CMOS sensor changed when the shutter time changed, which means that the change relationship of the grayscale value of each pixel must be found at different shutter times. Since all the data has to be gotten, using all of those data to calculate the gain value and offset of all pixels at different shutter time, too much data is needed. If calibration is required at another shutter time. The analysis must be re-examined. In that case, calibrating the data in a single shutter time become very inefficient. In Chapter 3, in addition to analyzing the data of a single shutter time, two methods are used to analyze the data at different shutter times. The first method is to find the set of relations between the gain value and the offset of different shutter times, and then calculate the gain value and offset at different shutter times. The second method is to find out the relationship between the grayscale value of each pixel and the shutter time, and calibrate the original data obtained at different shutter times by the two methods, which mentioned above. Finally compare both calibration result to the data, which calibrated by single shutter time method. In that case, we can confirm whether the two calibration methods are workable at different shutter times.