Auto Exposure And Auto Focusing Algorithm Research Base On Image Process

With the rapid development of electronics and information,the digital cameras have totally replaced the film cameras.The auto exposure and auto focusing are essential procedures in image quality evaluation,and also are important components in camera software.This dissertation focuses on the research of automatic exposure and auto focus,and reveals that it plays an important role in improving the quality of actual camera image and has practical significance for debugging the camera effect of the mobile phone.On one hand,as the auto exposure cannot determine the defect of the screen brightness level,this dissertation presents an automatic exposure algorithm based on the combination of average gray level and luminance histogram distribution.Firstly,we defined five gray intervals to determine whether the average brightness falls into the best gray interval.Secondly we calculated the remaining four brighter or darker areas to check if current luminance falls into special light environment.Finally by adjusting signal gain to make dynamic range of the gray,we improved the brightness level distribution.Experiments were given to evaluate the proposed algorithm,which is shown to meet the best gray interval and brightness histogram distribution.On the other hand,due to the complication of the image,there are some disadvantages exist in the current technology of auto focusing.How to improve the quality of the focused image and how to choose the focusing evaluation function for the focus area are yet to be solved.This dissertation firstly introduced the traditional sharpness evaluation algorithms.As the traditional algorithms have disadvantages such as low sensitivity and large amount of computation,in this dissertation we proposed an improved sharpness evaluation algorithm by using Tenengrad function.Based on the Sobel operator,the four-direction edge detection is improved.The simulation experiments are carried out by using five different edge images.The improved sharpness evaluation algorithm is shown to have high sensitivity,single-peak property and high stability.Then we analyzed the focusing window methods,through the comparison of sharpness evaluation,a better selecting method for area selection is proposed,by cutting the screen into a 3×3 block and putting the center area as the focus window.Experiments showed that it has the properties of high sensitivity and low time consumption.The auto exposure algorithm proposed in this dissertation not only can achieve best visual gray request,but also can make the image have better brightness level.The proposed auto focusing algorithm using four-direction Sobel operator has properties of high sensitivity,single-peak and high stability.The proposed auto focusing window has high sensitivity and short time consuming.The research in this dissertation is helpful for the future research on high-dynamic-range scene exposure and complex texture scenes for fast and accurate focusing.