A Method Of Measuring The Parameters Of Image Sensor By Particle Swarm Optimization

Image sensor is a device that converts optical signal into electronic signal,which is the core component of image acquisition system.In recent years due to CMOS image sensor in the digital camera,smart phones,security surveillance cameras,security and security related to the new weapons,and even the highest embodiment of human science and technology development-space science,have image sensor figure.With the CMOS image sensor applications more and more widely,the image sensor performance testing,evaluation and screening to ensure product quality and application reliability is very necessary.At present,the international research and testing of image sensors are commonly used in the "European Machine Vision Association" to develop the "1288 standard"(EMVA1288),the study based on particle swarm optimization and photon conversion theory,for image sensor several main performance parameters:Gain k,dark current,the initial dark signal mean,put forward a new test method.The test method of this paper is designed by EMVA1288 standard and particle swarm optimization algorithm,which can be obtained by inverting the system gain,dark signal mean and dark signal initial value of image sensor,and using the bright field and dark field gray Value of the average data,so that the test results more accurate.This paper first summarizes the development history and practical application of image sensor,and then elaborates the test standard of image sensor and the components of test system.The paper mainly introduces several key parameters of image sensor.The test data of EMVA1288 standard is obtained by using a set of image sensor test system,including the mean and variance of bright field and dark field.Based on the EMVA1288 standard,a mathematical model is obtained.By using the fitness function in the improved particle swarm algorithm,the mathematical model and the algorithm are combined to propose a new image sensor test method.In this method,the inertia weight of the particle swarm algorithm and the control of the late optimization process of the algorithm make the search precision of the algorithm more accurate.The experimental results show that this method works well.