发动机小直径管X射线底片数字化图像增强技术  

Digital Image Enhancement on X-ray Film of Aeroengine Small Diameter Tube

作　　者：王克广 岳乔 缪伟 邬冠华[2] WANG Ke-guang;YUE Qiao;MIAO Wei;WU Guan-hua(AECC Aviation Power Co.,Ltd.,Xi’an 710000,China;School of Testing and Optoelectronic Engineering,Nanchang Hangkong University,Nanchang 330063,China)

机构地区：[1]中国航发动力股份有限公司,西安710000 [2]南昌航空大学测试与光电工程学院,南昌330063

出　　处：《失效分析与预防》2022年第2期86-95,共10页Failure Analysis and Prevention

基　　金：国家自然科学基金(62161030)。

摘　　要：针对X射线胶片扫描后图像退化问题,提出一种基于麻雀搜索的改进多尺度Retinex(Multi-Scale Retinex,MSR)增强算法。以航空发动机小直径管数字化图像为研究对象,把图像梯度、频域特性和信息熵作为麻雀搜索算法的种群适应度函数,寻找双边滤波器中关键参数σ;最优值,实现自适应图像增强。两类典型实际图像测试结果表明,相比限制对比度直方图均衡化和传统MSR处理方法,本研究算法不仅在主观图像上能更加清楚地显示缺陷,而且在客观评价指标上至少提高10%以上,明显改善图像质量。Aiming at the problem of image degradation after X-ray film scanning,an improved multi-scale retinex(MSR)enhancement algorithm based on sparrow search is proposed.Taking the digital image of aeroengine small diameter tubes as the research object,the image gradient,frequency domain,and information entropy are employed as the population fitness function of the sparrow search algorithm.Then,realize the adaptive image enhancement is found the optimal values of key parameters in the bilateral filter.Based on the above methods,two kinds of typical actual images are recorded.Compared with limited contrast histogram equalization and traditional MSR processing methods,the proposed method can show defects more clearly on subjective images,and the objective evaluations are improved by at least 10%.The method proposed in this paper can significantly improve the image quality.

关 键 词：X射线底片扫描 图像增强 改进MSR 麻雀搜索算法 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]