光照不均管道内图像增强算法的研究与应用  

Research and Application of Image Enhancement Algorithm in Uneven Brightness Pipeline

作　　者：张建军[1,2] 胡惠灵 刘征宇[2] 解新胜[1] 

机构地区：[1]安全关键工业测控技术教育部工程研究中心,合肥230009 [2]合肥工业大学计算机与信息学院,合肥230009

出　　处：《计算机工程》2011年第16期227-229,共3页Computer Engineering

基　　金：广东省教育部产学研结合基金资助项目(2009B090300302)

摘　　要：采用机器视觉技术对大型引水压力钢管内的裂纹进行检测,容易受现场环境限制,使摄像机获取到的图片存在光照不均匀和掺杂噪声等问题。若不进行有效的预处理来消除非均匀背景光和噪声,会影响后期裂纹检测和提取结果。为此,提出一种新型图像增强算法,该方法基于线性空间滤波原理对图片的背景光进行拟合,以改善光照不均匀现象;利用小波变换去除噪声增强图像细节。给出常见的增强算法进行比较分析。实验结果表明,该方法优于传统的图像增强方法且运算量更小,可以满足实时处理的需求。The limitation of actual environment causes the problem of uneven brightness and noise in the image when capturing images of the crack detection by employing the technology of machine vision in large penstock.The results of crack detection and extraction are seriously affected if the uneven brightness and noise are not dealt with properly.To settle this problem,a new image enhancement method based on linear spatial filter and wavelet transformation is proposed.This method uses the linear spatial filter to improve the uneven brightness phenomenon.The wavelet transformation is employed to remove the noise and enhance the details.The comparison and analysis of several methods are presented.Experimental results show the method is superior to traditional image enhancement methods and is suitable for real-time processing.

关 键 词：裂纹检测 光照不均 线性空间滤波 小波 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]