基于改进小波变换的航空磁性零件去噪方法研究  

Research on Denoising Method of Aeromagnetic Parts Based on Improved Wavelet Transform

作　　者：陈琳琳 邓华军[1] 杜传红[1] 张殿喜[1] 陈召松[1] CHEN Linlin;DENG Huajun;DU Chuanhong

机构地区：[1]安顺学院,贵州安顺561000

出　　处：《科技创新与应用》2023年第25期11-14,共4页Technology Innovation and Application

基　　金：贵州省教育厅青年科技人才成长项目(黔教合KY字[2020]141)。

摘　　要：航空零件的质量控制是其工业生产的核心环节,基于计算机视觉的缺陷检测技术是实现航空零件质量控制的关键手段。受环境光照、成像设备等的影响,通过工业相机获取的航空零件图像会受到噪声干扰,信噪比较低,噪声污染会极大程度地降低图像质量,干扰后续缺陷检测的表现。因此,构造一种在有效去除噪声的同时又不会模糊边缘及细节的去噪方法是航空磁性零件缺陷检测的前提条件。该文将同源图像的小波系数融合与小波阈值去噪相结合,提出一种改进的小波变换方法并将其用于磁性零件去噪,相比硬阈值、软阈值及中值滤波方法,该文方法的去噪图像在视觉效果上更清晰,且性能指标更好。The quality control of aviation parts is the core of its industrial production.The defect detection technology based on computer vision is the key means to accomplish the quality control of aviation parts.Due to the influence of environmental illumination and imaging equipment,the images of aviation parts acquired by industrial cameras would be disturbed by noise with low signal-to-noise ratio.Noise pollution will greatly reduce the image quality and interfere with the performance of subsequent defect detection.Therefore,the construction of a denoising method which can effectively remove noise without blurring edges and details is a prerequisite for defect detection of aeromagnetic parts.In this paper,with the combination of wavelet coefficient fusion and wavelet threshold denoising of homologous images,an improved wavelet transform method is proposed and applied to magnetic parts denoising.Compared with hard threshold,soft threshold and median filtering methods,the denoising image of this method is clearer in visual effect and better in performance.

关 键 词：小波分解 系数融合 阈值 缺陷检测 去噪方法 

分 类 号：TN957.52[电子电信—信号与信息处理]