A Study Of 3A Algorithm Based On Image Analysis

With the significant improvement on the performance of hardware like microprocessors and image sensors,how to get images with higher quality becomes widespread concerned.In order to obtain a high-quality image,not only a good performed hardware is needed,a better image preprocessing algorithm is necessary too.3A algorithm,including automatic white balance algorithm(AWB),auto-focusing algorithm(AF),automatic exposure algorithm(AE),is an important part of image pre-processing algorithm,which can control sensors device based on image information.In this paper,many of the classic 3A algorithm is studied through analysis the characteristics and application.This paper introduces a kind of 3A method combining the advantages of different algorithm,which is used to deal with the real-time and multi-applicability image processing.Firstly,this paper describes the basic concepts of image processing,including the knowledge of image color science,imaging theory,the image acquisition model,and the general steps of digital image process.Secondly,the paper describes the basic process of AE control,AF control and AE control.Several existing classical AWB algorithms are analyzed,such as gray world algorithm,Gray World-Retinex algorithm,fuzzy rule method and an automatic white balance method based on edge detection,and an AWB algorithm based on dynamic threshold is proposed;the AE algorithm proposed based on lighting-condition detection,the detail of the design and implement on the algorithm are also given,the different clarity evaluation functions such as gray gradient function,frequency domain function and informatics entropy function are compared.Image search algorithms and the focus window selection part are also the necessary steps of AF control.Then some Simulations are done under different environment.Images captured in day and night,indoor and outdoor and those obtained from Internet are all been test.Objective and subjective evaluation based on time complexity,brightness and color value are used to judge the results of each experiment.Finally AWB and AE method combine together to process the image which is at underexposure and color distortion condition.The simulations turned out that,the AWB algorithm based on dynamic threshold can correct the color distortion under different lighting conditions,both the image with single color and rich colors.Especially when the algorithm is used on images captured indoor and having single color,the result turned out to be very satisfied,at the same time,the performance on images outdoor with single color have a big advancement.The AE algorithm based on lighting-condition detection perform very well at underexposure condition and slight overexposure condition.AF control will performance better by connecting several different clarity evaluation functions and using advanced Image search algorithms.So the 3 A algorithms used in this paper can meet the requirements of improving the image quality.Obviously,the image becomes more nature,clearer and brightness suitable.