基于灰轴调整的彩色图像自动白平衡  

Gray-Axis Alignment Based Automatic White Balance for Color Images

作　　者：王建[1] 庞彦伟[1] 俞浩[2] 

机构地区：[1]天津大学电子信息工程学院,天津300072 [2]天津大学理学院,天津300072

出　　处：《天津大学学报》2010年第6期479-484,共6页Journal of Tianjin University(Science and Technology)

基　　金：国家自然科学基金资助项目(60975001)

摘　　要：自动白平衡是一种重要的偏色校正技术.提出了一种新的基于灰轴调整的自动白平衡算法.首先引入理想灰轴和估计灰轴的概念.该方法基于白斑假设,首先选取图像中的部分像素点计算估计灰轴,然后借助旋转变换和尺度变换,将估计灰轴调整至与理想灰轴重合,以此实现彩色图像白平衡处理.实验结果表明,与经典的白斑法和灰色世界法相比,所提算法能有效校正均匀光照条件下的偏色图像.Automatic white balance is an important techmque for color cast correcuon ano a novel automauc wnlte balance algorithm based on gray-axis alignment has been proposed. The concepts of ideal gray axis and estimated gray axis are introduced. On the basis of white patch hypothesis, some pixels were first extracted from the image to compute the estimated gray axis. Then rotation transformation and scale transformation were applied for alignment of estimated gray-axis with ideal gray-axis so that white balance was implemented for color images. Experimental results show that compared with classical white patch method and gray world method,the proposed method is effective in color cast correction under uniform illumination.

关 键 词：偏色校正 自动白平衡 灰轴调整 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]