一种基于图像分区的自动曝光算法  

An automatic exposure algorithm using scene analysis

作　　者：甘玉泉[1,2] 高伟[1] 

机构地区：[1]中国科学院西安光学精密机械研究所,陕西西安710119 [2]中国科学院研究生院,北京100039

出　　处：《微计算机信息》2009年第12期303-304,273,共3页Control & Automation

摘　　要：自动曝光控制系统是现代数码相机和摄像机的不可缺少的模块之一。提出一种基于图像分区的自动曝光控制算法。当拍摄对象处于背光或者正面强光时,自动曝光也能很好地控制。根据人类的视觉机制,可认为中间区域为目标区域,背景区域为不感兴趣区域,图像分区即将拍摄场景划分为几个区域,并对每一个区域赋予不同的权值,从而计算图像的加权亮度均值。Automatic exposure control （AEC） system is one of the indispensable parts of modern digital still camera and video camera. An automatic exposure algorithm using scene analysis is introduced. Auto exposure is required to perform well even in special conditions such as backlighting and excessive front-lighting. According to the attention mechanism of human visual system （HVS）, it is considered that the middle part of the scene is the object, the background is uninterested part. Scene analysis involves segmenting the scene into different segments and weighting each segment appropriately to calculate an exposure to match the application question. And the mean weighted luminance of the entire image is calculated.

关 键 词：自动曝光控制 图像分区 加权亮度均值 

分 类 号：TP391.72[自动化与计算机技术—计算机应用技术]