Research On Crosstalk Characteristics Of Adjacent Pixel Based On Potentital Difference

At present,image sensors are widely used in important fields such as civil and military.With the rapid development of image sensors integrated manufacturing process technology,the pixel unit size and the distance between pixel units of the image sensors have become smaller and smaller,the probability of pixel crosstalk has been greatly increased.The existence of pixel crosstalk will reduce the quality of the output image in the image sensor imaging system.Therefore,the research on the mechanism,measurement and compensation algorithm of crosstalk is of great significance.At present,domestic and foreign research on solving the problem of pixel crosstalk of image sensors mainly focuses on the following two aspects:Firstly,the crosstalk can be reduced to a reasonable level by improving the structure and materials;Secondly,crosstalk noise is compensated by adding crosstalk compensation algorithm in image processing.In this thesis,an algorithm compensation method based on potential well crosstalk was proposed by describing,measuring,modeling crosstalk and so on.This thesis introduces the research progress of image sensor pixel crosstalk at home and abroad,and explains the basic working principle of CCD image sensor and the basic structure of MOS capacitor.Based on the semiconductor working principle and pixel potential well model,the source,mechanism and diffusion model of image sensor pixel crosstalk were analyzed.Designed a test system solution for image sensor pixel crosstalk,that is,collecting crosstalk data under different combinations of exposures of adjacent pixels in the pixel array.The crosstalk characteristic is realized in the case of different combinations of exposure intensity of each pixel in the adjacent pixel array.Through the computer simulation of the crosstalk diffusion model under different parameter conditions,the crosstalk data is obtained by using the simulation to output the crosstalk data,and the curve relationship of the crosstalk coefficient is obtained.Two types of image sensor algorithm compensation models based on the linear array image sensor "one" model and the area array image sensor "cross" model based on the potential difference between adjacent pixel potential wells are proposed.Related to the ideal gray value,the theoretical gray value of the pixel after crosstalk compensation of the image sensor is obtained through matrix inversion.The simulation crosstalk data under different exposure conditions is used to verify the crosstalk compensation algorithm model.The results show that the central pixel unit can compensate the crosstalk component of its crosstalk to the adjacent location unit.The difference between the crosstalk gray-scale component compensated by the contrast central pixel unit and its theoretical gray-scale value is between 2-8,and the crosstalk gray-scale component compensated by the adjacent location pixel unit is suppressed within 0-3 among them,the crosstalk area is basically effectively suppressed.