基于机器视觉的噪声处理技术研究  

Research on Noise Processing Technology Based on Machine Vision

作　　者：翁蒙婷 WENG Mengting(State Grid Zhejiang Electric Power Co.,Ltd.,Taizhou Luqiao District Power Supply Company,Taizhou 317000,China)

机构地区：[1]国网浙江省电力有限公司台州市路桥区供电公司,浙江台州317000

出　　处：《电声技术》2024年第11期10-12,共3页Audio Engineering

摘　　要：针对配电网建设中的图像噪声问题,提出一种基于机器视觉的多层次噪声抑制技术,旨在改善配电设备图像的质量。该技术结合自适应直方图均衡化、深度学习卷积神经网络与多尺度特征融合技术,解决配电设备图像中的噪声问题。实验结果显示,与传统去噪算法相比,所提技术在峰值信噪比(Peak Signal-to-Noise Ratio,PSNR)和结构相似度(Structural Similarity,SSIM)方面表现优异,能够显著提升图像处理质量。A multi-level noise suppression technology based on machine vision to address the issue of image noise in the construction of distribution networks,aiming to improve the quality of images of distribution equipment.The technology combines adaptive histogram equalization,deep learning convolutional neural networks,and multi-scale feature fusion techniques to solve the noise problem in distribution equipment images.Experimental results show that compared with traditional denoising algorithms,the proposed technology performs excellently in terms of Peak Signal-to-Noise Ratio(PSNR)and Structural Similarity(SSIM),significantly enhancing the quality of image processing.

关 键 词：机器视觉 噪声处理 深度学习 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]