Research On Image Sensor Correction Method Based On Pixel Photon Transfer Characteristics

At present,the integration process of image sensors is very mature,and ultra-high speed and ultra-high resolution image sensors have been extensively used in many fields.However,due to the limitation of the manufacturing process,the parameter characteristics of each pixel of the image sensor are different,as the result,the signals output of each pixels are inconsistent under the exposure of the same condition.This inconsistency leads to the unevenness of the output image.Secondly,the photon transfer curve of the image sensor shows nonlinearity with the increase of exposure time.As results of this nonlinear phenomenon,the dynamic range of the image sensor is reduced,as well as makes the contrast of the output image worse.Therefore,it is necessary in order to correct the image sensor.The integration process of image sensors is very mature,making ultra-high-speed,ultra-high resolution image sensors widely used in many fields in recent years.However,due to the limitation of the manufacturing process,the parameter characteristics of each pixel of the image sensor are different,as well as resulting in inconsistencies in the signals output by the respective pixels under the exposure of the same condition,and such inconsistency will result in unevenness of the output image.This inconsistency leads to the unevenness of the output image.At the same time,as the exposure time increases,the photon conversion curve of the image sensor will exhibit nonlinearity.This nonlinear phenomenon will eventually reduce the dynamic range of the image sensor,resulting in poor contrast of the output image.Therefore,it is necessary to perform correction processing on the image sensor.This paper summarizes the research progress of image sensor calibration methods in various countries,and discusses and compares the existing calibration methods.Aiming at the problem that the traditional image sensor correction method is only a single image sensor non-uniformity or nonlinearity correction and poor versatility,an image sensor calibration method based on pixel photon conversion is proposed based on photon conversion theory.This method combines the photon conversion theory to analyze the characteristics and sources of various types of noise of image sensors.By fitting the function expression of the gray-scale response curve of each pixel of the image sensor,the correction conversion matrix of the image sensor is calculated,and then the image is exposed.The original image gray matrix is brought into the gray scale response curve of the corresponding pixel to inverse the exposure characterization matrix.Finally,an ideal target image sensor photon conversion curve function relationship is assumed,and the exposure characterization matrix is brought into the corrected image gradation matrix,that is,the corrected image.In addition,in order to make the CCD image sensor work properly and provide necessary experimental data,this paper builds an image sensor test system,which can provide stable light and dark field uniform exposure test conditions for the image sensor and isolate the interference of external environment light.On the basis of the original image sensor testing system,a CCD image acquisition circuit based on the timing control of FPGA is designed.The circuit consists of a power management circuit,timing the drive circuit,video signal conditioning circuit and analog-to-digital transfer circuit.Practical test shows that the system works stably and meets the test requirements of the CCD image sensor.At the same time,a special correction software is compiled with MATLAB to fit the gray response curve function expression of each pixel,and the correction transfer matrix of the CCD image sensor is obtained.Then the correction transfer matrix is saved to the FPGA to correct the output gray value of the image sensor.Finally,a scientific-level CCD and a scientific-level CMOS image sensor are used as experimental objects.The pixel response curves and image quality of the image sensor before and after correction are compared.The experimental results show that the non-linearity and non-uniformity of the image sensor are effectively suppressed,the contrast and uniformity of the output image are significantly improved,and the insensitive pixels of the image sensor can be corrected.In addition,the correction effect of the algorithm on CCD image sensor and CMOS image sensor indicates that it has obvious versatility.