面向小目标测量的通道注意力网络与系统设计  

Design of channel attention network and system for micro target measurement

作　　者：傅扬伟 张进[1,2,3] 孙珍惜 张瑞[1,2] 李维诗 夏豪杰 FU Yangwei;ZHANG Jin;SUN Zhenxi;ZHANG Rui;LI Weishi;XIA Haojie(School of Instrument Science and Opto-electronics Engineering,Hefei University of Technology,Hefei 230009,China;Anhui Province Key Laboratory of Measuring Theory and Precision Instrument,Hefei 230009,China;Engineering Research Center of Safety Critical Industrial Measurement and Control Technology,Ministry of Education,Hefei 230009,China)

机构地区：[1]合肥工业大学仪器科学与光电工程学院,安徽合肥230009 [2]测量理论与精密仪器安徽省重点实验室,安徽合肥230009 [3]教育部安全关键工业测控技术工程研究中心,安徽合肥230009

出　　处：《光学精密工程》2023年第6期962-973,共12页Optics and Precision Engineering

基　　金：国家自然科学基金面上项目资助(No.52175504,No.51927811);中央高校基础研究基金资助项目(No.PA2021KCPY0027,No.PA2021GDGP0061)。

摘　　要：微器件广泛应用于电子工业。由于衍射效应,微器件的物理边缘与光学边缘不一致,这给检测和测量带来了挑战。为提高微目标检测与测量精度,本文将图像超分辨率重建与目标测量结合,提出了一种基于边缘增强的图像超分辨率重建算法并搭建了对应的测量系统。首先提出了一种新的图像超分辨率重建质量评价参数,证明了图像超分辨率重建提高目标测量精度的可行性。针对目标边缘,将通道注意力机制引入网络,增强了网络对图像边缘的重建能力。最后,设计并搭建了目标测量系统,并进行了实验。结果表明:在公开数据集上,本文算法能取得更高的峰值信噪比(PSNR)和结构相似性(SSIM)等客观指标值;在实际测量中,本文算法可将原有测量系统极限分辨率提高25.9%,目标测量精度平均提高51.6%。本文研究为工业生产中的微目标检测和测量提供了一个潜在的发展方向。Microdevices are widely used in the electronic industry.However,the diffraction effect,which causes misalignments in the physical and optical edges of micro devices,brings challenges to detection and measurement.To address this issue,this study combines image super-resolution reconstruction with tar⁃get measurement to propose an image super-resolution reconstruction algorithm based on edge enhance⁃ment and build a corresponding measurement system.In this study,a new quality evaluation parameter is proposed for image super-resolution reconstruction,to prove the feasibility of super-resolution reconstruc⁃tion in improving target measurement accuracy.Aiming at the target edge,a channel attention mechanism is also introduced into the network to enhance its ability to reconstruct the image edge.Finally,the target measurement system is designed and built,and experiments are carried out.The results show that the pro⁃posed algorithm can achieve higher peak signal to noise ratio(PSNR)and structural similarity(SSIM)values on an open dataset.In real-world measurements,this algorithm improved the limit resolution of the original measurement system by 25.9%and the target measurement accuracy by 51.6%,on average.This study provides a potential direction for the development of micro-target detection and measurement in industrial production.

关 键 词：深度学习 图像超分辨率 计算机视觉 边缘检测 亚像素 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]