基于高斯频域低通滤波和图像差分的字符图像光照不均校正法  

Character images uneven illumination correction method based on Gaussian frequency low pass filtering and image difference

作　　者：杨颖[1] 戴彬[2] 

机构地区：[1]川北医学院基础医学院,四川南充637100 [2]西华师范大学商学院,四川南充637009

出　　处：《华中师范大学学报（自然科学版）》2013年第2期181-184,共4页Journal of Central China Normal University：Natural Sciences

基　　金：四川省教育厅2010年青年基金项目(10ZC102)

摘　　要：光照不匀会对字符图像检测带来极大的负面影响.基于高斯频域低通滤波和图像差分提出一个新的字符图像光照不均校正法,该方法首先根据照明光场和图像细节分别对应低频分量和高频分量的特点,设计高斯低通滤波器去除光照背景以获取照明光场,然后再以此为根据对字符图像进行校正.同时,针对高斯低通滤波器参数选择上的难点,提出了实验法确定参数σ加逐次滤波对照明光场进行逼近的方法,以确保最终校正结果的准确.实验结果表明该方法校正效果较好.Uneven illumination can result in serious negative impact on character image detection. Based on Gaussian frequency low pass filtering and image difference, a new character image uneven illumination correction method was proposed. According to the characteristics which Lighting light field and image detail respectively corresponding to low-frequency components and high-frequency components, gaussian frequency low pass filtering was firstly designed to remove light background which aims to get lighting light field. And then, character image was corrected based on the designed gaussian frequency low pass filtering. Meanwhile, for the difficulty on choice of gaussian low-pass filter parameter, a method which obtains parameter 6 based on experimentation and successive filtering on lighting light field was presented to ensure the end correction results. The experimental results show that the correction effect of the proposed method is better.

关 键 词：图像检测 高斯频域低通滤波 照明光场 低频分量 高频分量 

分 类 号：TP242.62[自动化与计算机技术—检测技术与自动化装置]