基于特征图规正的虹膜图像自适应Gamma校正方法  

Adaptive Gamma correction method of Iris image based on characteristic pattern

作　　者：马琳[1] 王钧慧[1] 王宽全[1] 张大鹏[2] 

机构地区：[1]哈尔滨工业大学计算机科学与技术学院,黑龙江哈尔滨150001 [2]香港理工大学计算机系

出　　处：《燕山大学学报》2010年第2期173-179,共7页Journal of Yanshan University

基　　金：黑龙江省归国人员科研基金项目(LC04C07);哈尔滨市后备人才计划基金项目(2003AFLXJ006)

摘　　要：在虹膜图像采集过程中产生的光斑使得整幅虹膜图像的照度不均匀。为了消除照度不均匀对特征提取的影响,本文提出了一种基于特征图规正的自适应Gamma校正方法。该方法将虹膜图像的特征图作为Gamma校正的依据,把距离因素考虑到Gamma校正中,而且利用模拟退火算法实现了关键值Gamma的自适应调整。该方法使Gamma值的变化更加趋于合理,提高了Gamma矫正对光照变化的适应能力,避免了较正后图像的失真。实验证明该方法不仅可以尽可能的提高图像的质量,而且对校正后的图像进行边缘提取,可以获得更加清晰连续的边缘。Flares are produced during image acquisition, treat illumination unequality on images, especially iris image and greatly reduce image quality. To eliminate such effect of illumination, in this paper, an adaptive Gamma correction method for iris images based on feature graph alignment is presented. This method takes characteristic patterns of iris images as basis for Gamma crrection, distance measure definition and distance calculation. Simulated annealing algorithm is applied for Gamma value estimation, thus more suitable Gamma values may be achieved. This method enhances adaptive capacity of Gamma correction and avoids image distortion. Experimental results show this method can not only enhance image quality, but can also obtain clearer edges through Canny operator edge detection.

关 键 词：虹膜图像 修正的Gamma校正 自适应 模拟退火 

分 类 号：TN911.73[电子电信—通信与信息系统]