Font Size: a A A

A Research Of Testing Technology For Main Noises Of CMOS Image Sensor

Posted on:2018-08-13Degree:MasterType:Thesis
Country:ChinaCandidate:C L MaoFull Text:PDF
GTID:2348330542491437Subject:Optical Engineering
Abstract/Summary:PDF Full Text Request
The integrated technology of image sensor has been very mature nowadays.Ultra-high speed and ultra-high resolution image sensor has been widely applied in many fields.Before the application on high-end areas like military and aerospace,testing the characteristics of image sensor is necessary to ensure the reliability of the project.The test for main parameters of image sensor is to analyze the relationship between signal and noise of the image sensor in general image sensor testing standard.Testing the main noise of image sensor reasonably and stably is the prerequisite for testing parameters of image sensor.The theory of Photon Transfer is a basic theory describing the connection between signal and noise of image sensor,as well as an important theoretical basis for image sensor testing.In this paper,the software and algorithm was designed mainly based on the theory of Photon Transfer and EMVA standard 1288.This paper measured trend of several kinds of noises stably,by means of analyzing output gray value data of image sensor and calculating the image noise under different exposure conditions.This paper outlined the research progress of image sensor testing domestic and international first,then elaborated the theory of photon transfer and the standard of CMOS image sensor testing,introduced noise of the image sensor and listed the equivalent noise model,put forward several test methods of main noise,including the test method of shot noise and dark current,and the suppression technique of FPN.This paper has designed the image sensor testing system and the test software,analyzed the output image data of the image sensor testing system tested the characteristics of shot noise and dark current,put forward effective FPN correction method,and compared the effect of linear correction and nonlinear flat field correction.This paper used the least square fitting algorithm to correct FPN,reduced nonlinear problem of the FPN by quadratic fitting.The correction method could correct FPN in arbitrary exposure time by the experiment,linear response of the corrected image was better.
Keywords/Search Tags:CMOS image sensor, Shot noise, Dark current, FPN, Flat field correction
PDF Full Text Request
Related items