复杂异形工件射线图像对比度自适应提升算法  

Adaptive enhancement algorithm for contrast of complicated shaped workpiece radiographic image

作　　者：李辉[1] 陈平[1] Li Hui;Chen Ping(Shanxi Province Key Laboratory of Information Detection and Processing,North University of China,Taiyuan 030051,China)

机构地区：[1]中北大学信息探测与处理山西省重点实验室,太原030051

出　　处：《国外电子测量技术》2021年第11期1-6,共6页Foreign Electronic Measurement Technology

基　　金：国家自然科学基金(61801437,61871351,61971381)项目资助。

摘　　要：针对复杂异型构件因其腔体结构复杂而导致的X射线图像的低对比度、细节信息不明显等问题,从图像自适应增强、特征显化方面展开了研究,提出了一种基于图像局部方差与变分法的复杂异型构件的射线图像对比度提升算法。首先,提取图像的对比度信息并在此基础上构建基于对比度场的图像对比度提升模型;其次,计算图像的局部方差,结合对比度场提升模型,利用图像的局部方差设计自适应增益函数对上述图像对比度提升模型进行改进,并采用;最后,在提升后的梯度场与目标图像之间建立最小化泛函方程,并通过变分法对方程进行求解,最终获取目标图像。实验结果表明,该算法可以较全面的提升图像对比度,增强图像细节信息,并压缩图像动态范围,从而得到高质量的图像,进一步提高无损检测的识别率。Aiming at the problems of low contrast and inconspicuous detail information of X-ray images caused by the complex structure of the cavity of complex shaped components, this paper carried out research from the aspects of image self-adaptive enhancement and feature visualization, and proposed a method based on image local variance Contrast enhancement algorithm of ray image of complex and special-shaped components with variational method. First, extract the contrast information of the image and build an image contrast enhancement model based on the contrast field;second, calculate the local variance of the image, combine the contrast field enhancement model, and use the local variance of the image to design an adaptive gain function for the above image contrast The promotion model is improved and adopted;finally, a minimization functional equation is established between the promoted gradient field and the target image, and the equation is solved by the variational method, and the target image is finally obtained. Experimental results show that the proposed algorithm can comprehensively improve image contrast, enhance image detail information, and compress image dynamic range, thereby obtaining high-quality images, and further improving the recognition rate of non-destructive testing.

关 键 词：X射线成像 复杂结构件 低对比度增强 局部方差 变分法 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] O434.1[自动化与计算机技术—计算机科学与技术]