DR image enhancement of aero⁃engine guide vane based on improved CLAHE
-
摘要:
为了解决航空发动机导向叶片数字射线（DR）检测图像信息动态范围大、对比度低、细节信息不明显，缺陷区域难以识别的问题，提出一种改进型限制对比度自适应直方图均衡化（CLAHE）算法。采用CLAHE增强导向叶片DR图像对比度，同时引入基于空间均值滤波器的Gaussian掩模处理，进行DR图像降噪，提取DR图像的低频信息；采用CLAHE增强的图像与提取的DR图像低频信息线性做差，突出DR图像的高频细节信息；与CLAHE增强的图像线性叠加，进一步提高了DR图像的对比度，实现导向叶片DR图像增强。依据图像基本空间分辨率（SRB）、信噪比（SNR）、灰度平均值对DR图像增强效果进行评价。结果表明: 改进的CLAHE算法，可以同时将表征SRB的D13双丝线对应的调制深度值从49.17%提高到了56.08%,整体灰度平均值从32 400.66增加到了38 684.43，02号微小裂纹缺陷的SNR从14.10提升到了15.16。结果显示优化的CLAHE算法，相比自适应直方图均衡化（AHE）等4种经典的航空发动机导向叶片DR图像增强算法，不仅提高了平坦区域对比度，突显了边缘细节信息，而且有效提升了微小缺陷的视觉效果。
-
关键词:
- 数字射线（DR）图像 /
- 微小缺陷 /
- 导向叶片 /
- 限制对比度自适应直方图均衡化（CLAHE） /
- 图像增强
Abstract:In order to solve the problems of large dynamic range，low contrast，inconspicuous detail information，and difficulty in identifying defect areas in the image information detected by digital radiography （DR） of aero⁃engine guide blades，an improved adaptive histogram equalization algorithm with limited contrast was proposed.Contrast limited adaptive histogram equalization （CLAHE） was used to enhance the contrast of the DR image of the guide vane.Meanwhile，the Gaussian mask processing based on the spatial mean filter was introduced to reduce the image noise and extract the low⁃frequency information of the DR image.The high⁃frequency information of DR image was highlighted by difference of CLAHE enhanced image and the low⁃frequency information of DR image.Furthermore，a linear superimposing of CLAHE enhanced image and highlighted image high frequency information was employed to further improve the contrast of DR image and realize the enhancement of the DR image of the guide vane.In addition，the DR image enhancement effect was evaluated based on the basic spatial resolution of the image （SRB），signal⁃to⁃noise ratio （SNR） and gray average.Experimental results showed that the improved CLAHE algorithm can simultaneously increase the modulation depth value characterizing SRB correspondingly to the D13 double⁃filament wire from 49.17% to 56.08%，and the overall grayscale average value from 32 400.66 to 38 684.43，and the SNR of the No.02 micro crack defect was improved from 14.10 to 15.16.The result showed that the improved CLAHE algorithm not only improved the flat area contrast and highlighted the edge detail information，but also effectively improved visual effect of small defects in comparison with four classic aero⁃engine guide vane DR image enhancement algorithms such as adaptive histogram equalization （AHE）.
-
图 4 航空发动机导向叶片原图和局部DR图像
Figure 4. Original image and partial DR image of aero engine guide vane
图 6 导向叶片DR缺陷图像子块及灰度直方图
Figure 6. Sub⁃block and gray histogram of the DR defect image of the guide vane
图 7 自适应直方图均衡化处理结果及灰度直方图
Figure 7. Adaptive histogram equalization processing results and gray histogram
图 8 裁剪分配插值运算结果及灰度直方图
Figure 8. Interpolation operation results and gray histogram of crop distribution
图 10 Gaussian处理导向叶片DR图像子块及灰度直方图
Figure 10. Gaussian processing guide blade DR image sub⁃block and gray histogram
图 11 导向叶片DR图像低频信息及灰度直方图
Figure 11. Low⁃frequency information and gray histogram of the DR image of the guide vane
图 12 导向叶片DR图像高频信息及灰度直方图
Figure 12. High⁃frequency information and gray histogram of the DR image of the guide vane
图 13 改进CLAHE处理结果和灰度直方图
Figure 13. Improved CLAHE processing results and gray histogram
图 16 Gaussian滤波处理的导向叶片DR缺陷图像
Figure 16. DR defect image of guide vane processed by Gaussian filter
图 17 Unsharp锐化处理的导向叶片DR缺陷图像
Figure 17. DR defect image of guide vane processed byUnsharp sharpening
图 19 改进CLAHE处理的导向叶片DR缺陷图像
Figure 19. DR defect image of guide vane processed by improved CLAHE
图 22 AHE处理的导向叶片DR图像的测试结果
Figure 22. Test results of DR images of guide blades processed by AHE
图 23 Gaussian滤波处理的导向叶片DR图像的测试结果
Figure 23. Test results of DR images of guide blades processed by Gaussian filtering
图 24 Unsharp锐化处理的导向叶片DR图像的测试结果
Figure 24. Test results of DR images of guide blades processed by Unsharp sharpening
图 25 CLAHE算法处理的导向叶片DR图像的测试结果
Figure 25. Test results of DR images of guide blades processed by CLAHE algorithm
图 26 改进CLAHE算法处理的导向叶片DR图像的测试结果
Figure 26. Test results of DR images of guide blades processed by improved CLAHE algorithm
图 27 不同处理方式导向叶片DR整体图像和02号缺陷信噪比对比结果
Figure 27. Comparison results of the overall image of the guide blade DR with different processing methods and the signal⁃to⁃noise ratio of No.02 defect
表 1 不同处理方式导向叶片DR图像整体信噪比SNR
Table 1. Overall signal⁃to⁃noise ratio SNR of the DR image of guide blade with different processing methods
图像种类 灰度平均值 灰度均方差 信噪比/dB DR原始图像 32 400.66 9 634.52 3.36 AHE处理 32 578.79 10 724.24 3.04 Gaussian滤波 32 400.66 9 633.61 3.36 Unsharp锐化 32 412.68 9 645.22 3.37 CLAHE处理 33 522.93 9 926.95 3.38 改进CLAHE处理 38 684.43 9 998.56 3.87 
下载: 导出CSV
表 2 不同处理方式02号缺陷的信噪比SNR
Table 2. Signal⁃to⁃noise ratio SNR of No.02 defect in different processing methods
图像种类 灰度平均值 灰度均方差 信噪比/dB DR原始图像 28 232.77 2 002.11 14.10 AHE处理 25 628.03 5 973.01 4.29 Gaussian滤波 28 232.81 2 001.42 14.10 Unsharp锐化 28 232.47 2 009.34 14.05 CLAHE处理 34 085.50 4 130.98 8.25 改进CLAHE处理 31 159.31 2 055.01 15.16
下载: 导出CSV
-
[1] 周莹.航空发动机涡轮叶片数字射线成像研究[D].南昌：南昌航空大学,2017.ZHOU Ying.Research on digital radiography of aero⁃engine turbine blades[D].Nanchang:Nanchang Hangkong University，2017.（in Chinese） [2] 赵俊皓.空心涡轮叶片残芯X射线检测技术[D].哈尔滨:哈尔滨工业大学,2018.ZHAO Junhao.X⁃ray detection technology of hollow turbine blade residual core[D].Harbin:Harbin Institute of Technology，2018.（in Chinese） [3] 王攀峰,赵书俊,王双玲,等.一种基于Retinex原理的DR图像增强改进算法[J].中国体视学与图像分析,2020,25（1）:57⁃64.WANG Panfeng，ZHAO Shujun，WANG Shuangling，et al.An improved DR image enhancement algorithm based on Retinex principle[J].Chinese Stereology and Image Analysis，2020，25（1）:57⁃64.（in Chinese） [4] BZIMMERMAN J，PIZER S，STAAB E，et al.An evaluation of the effectiveness of adaptive histogram equalization for contrast enhancement[J].IEEE Transactions on Medical Imaging，1988,7（4）:304⁃312. [5] REZA A M.Realization of the contrast limited adaptive histogram equalization （CLAHE） for real⁃time image enhancement[J].Journal of Vlsi Signal Processing Systems for Signal Image and Video Technology，2004，38（1）:35⁃44. [6] 陶佩,付青青,吴爱平.一种限制对比度的快速POSHE图像增强算法[J].长江大学学报（自然科学版）,2020,17（2）:104⁃111.TAO Pei，FU Qingqing，WU Aiping.A fast POSHE image enhancement algorithm with limited contrast[J].Journal of Yangtze University （Natural Science Edition），2020，17（2）:104⁃111.（in Chinese） [7] 刘轩,刘佳宾.基于对比度受限自适应直方图均衡的乳腺图像增强[J].计算机工程与应用,2008（10）:173⁃175.LIU Xuan，LIU Jiabin.Breast image enhancement based on contrast⁃constrained adaptive histogram equalization[J].Computer Engineering and Applications，2008（10）:173⁃175.（in Chinese） [8] LOGANATHAN R，KUMARASWAMY Y S.Medical image compression with lossless region of interest using adaptive active contour[J].Journal of Computer ence，2012，8（5）:747⁃751. [9] KIM M ，CHUNG M.Recursively separated and weighted histogram equalization for brightness preservation and contrast enhancement[J].IEEE Transactions on Consumer Electronics，2008，54（3）:1389⁃1397. [10] 孙冬梅,陆剑锋,张善卿.一种改进CLAHE算法在医学试纸条图像增强中的应用[J].中国生物医学工程学报,2016,35（4）:502⁃506.SUN Dongmei，LU Jianfeng，ZHANG Shanqing.Application of an improved CLAHE algorithm in image enhancement of medical test strips[J].Chinese Journal of Biomedical Engineering，2016，35（4）:502⁃506.（in Chinese） [11] 王红茹,李瑞,王佳.基于改进CLAHE的水下彩色图像增强算法[J].舰船电子工程,2019,39（11）:119⁃122,141.WANG Hongru，LI Rui，WANG Jia.An underwater color image enhancement algorithm based on improved CLAHE[J].Ship Electronic Engineering，2019，39（11）:119⁃122，141.（in Chinese） [12] 张文霞,王春光,王海超,等.改进的CLAHE无芒隐子草叶切片图像增强[J].农机化研究,2022,44（1）:25⁃30.ZHANG Wenxia，WANG Chunguang，WANG Haichao，et al.Improved image enhancement of CLAHE without awn cryptocarp leaf slices[J].Journal of Agricultural Mechanization Research，2022，44（1）:25⁃30.（in Chinese） [13] 刘陈瑶,胡梦竹,张龙飞.基于改进CLAHE的SF_6红外图像增强[J].光学技术,2021,47（1）:107⁃112.LIU Chenyao，HU Mengzhu，ZHANG Longfei.SF_6 infrared image enhancement based on improved CLAHE[J].Optics Technology，2021，47（1）:107⁃112.（in Chinese） [14] İLK H G，JANE O ，İLK Ö.The effect of Laplacian filter in adaptive unsharp masking for infrared image enhancement[J].Infrared Physics and Technology，2011，54（5）:427⁃438. [15] 徐艳丽,谭剑.基于非线性反锐化掩膜的DR图像边缘增强[J].中国医学物理学杂志,2010,27（4）:2013⁃2015.XU Yanli，TAN Jian.DR image edge enhancement based on nonlinear unsharp masking[J].Chinese Journal of Medical Physics，2010，27（4）:2013⁃2015.（in Chinese） [16] RAMPONI G.A cubic unsharp masking technique for contrast enhancement[J].Signal Processing，1998，67（2）:211⁃222. [17] 黄勇.基于双边滤波和改进CLAHE算法的低照度图像增强研究[D].湖南 湘潭:湘潭大学,2019.HUANG Yong.Research on low⁃light image enhancement based on bilateral filtering and improved CLAHE algorithm[D].Xiangtan Hunan：Xiangtan University，2019.（in Chinese） [18] 许蓉,王直,宗涛.基于改进高斯滤波的医学图像边缘增强[J].信息技术,2020,44（4）:75⁃78.XU Rong，WANG Zhi，ZONG Tao.Medical image edge enhancement based on improved Gaussian filtering[J].Information Technology，2020，44（4）:75⁃78.（in Chinese） [19] 刘凯,王晓勇,袁孟春.碳纤维缠绕发动机壳体数字射线DR成像检测图像质量控制[J].宇航材料工艺,2020,50（5）:76⁃82.LIU Kai，WANG Xiaoyong，YUAN Mengchun.Image quality control of digital radiographic DR imaging inspection of carbon fiber wound engine shell[J].Aerospace Materials Technology，2020，50（5）:76⁃82.（in Chinese） -
点击查看大图
计量
- 文章访问数: 527
- HTML浏览量: 210
- PDF量: 38
- 被引次数: 0