基于小波变换的W18Cr4V图像去噪算法研究  

Research on W18Cr4V Image Denoising Algorithm Based on Wavelet Transform

作　　者：何英杰 石秀东 陈昊 张文利 HE Yingjie;SHI Xiudong;CHEN Hao;ZHANG Wenli(School of Mechanical Engineering,Jiangnan University,Wuxi,Jiangsu 214122,China)

机构地区：[1]江南大学机械工程学院,江苏无锡214122

出　　处：《轻工机械》2023年第1期59-65,共7页Light Industry Machinery

摘　　要：高速工具钢W18Cr4V退火状态下采集的图像内含有许多噪声,为解决基于传统的软、硬阈值函数在W18Cr4V图像去噪过程中出现图像失真和伪吉布斯现象等问题,课题组提出一种新的连续、可导、低偏差的阈值函数和自适应调整阈值的小波去噪算法。新的阈值函数不仅改善了软阈值函数去噪过程中由于固定偏差导致的图像失真问题,而且避免了由于硬阈值图像不连续产生的附加振荡现象;基于MATLAB环境,对3张W18Cr4V退火图像进行去噪实验。结果表明:与传统阈值函数相比,改进的阈值函数对退火图像处理后所得峰值信噪比较高、结构相似性较好、去噪效果较好且自适应能力较强。The image of high speed tool steel W18Cr4V annealed contains a lot of noise. In order to solve the problems of image distortion and pseudo-Gibbs phenomenon in the process of W18Cr4V image denoising caused by traditional soft and hard threshold functions, a new continuous, differentiable and low deviation threshold function and adaptive threshold wavelet denoising algorithm was proposed. The image distortion caused by fixed deviation in the denoising process of soft threshold function was improved by the new threshold function and the additional oscillation caused by hard threshold image discontinuity was avoided. Based on MATLAB, three W18Cr4V annealing images were denoised. The results show that compared with the traditional threshold function, the improved threshold function has higher peak SNR, better structural similarity, better denoising effect and better adaptive ability for annealing image processing.

关 键 词：高速工具钢 小波去噪 阈值函数 自适应阈值 峰值信噪比 结构相似性 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术] TH142.2[自动化与计算机技术—计算机科学与技术]