基于机器视觉的砂型铸件表面粗糙度检测研究  

Surface Roughness Detection of Sand Casting Based on Machine Vision

作　　者：贾栋贤 李广宇[1,2] 刘文文 付宏源 姚山 IA Dongxian;LI Guangyu;LIU Wenwen;FU Hongyuan;YAO Shan(School of Materials Science and Engineering,Dalian University of Technology,Dalian 116024;Ningbo Institute of Dalian University of Technology,Ningbo 315016)

机构地区：[1]大连理工大学材料科学与工程学院,大连116024 [2]大连理工大学宁波研究院,宁波315016

出　　处：《特种铸造及有色合金》2025年第1期23-28,共6页Special Casting & Nonferrous Alloys

基　　金：国家自然科学基金资助项目(52205359);中央高校基本科研业务费专项基金资助项目[DUT23RC(3)039]。

摘　　要：结合图像处理技术,使用铸造表面粗糙度标准样块的表面图像对融入注意力机制的残差卷积神经网络进行了训练,并以此为核心开发了检测软件,配合图像采集模块用于实体铸件的检测,实现了铸件表面粗糙度的非接触测量。实测铸件的准确率为87.5%,单次测量用时0.086 s,能够在实际生产中快速、便捷地提供测量结果。Surface images of blocks with standard roughness was utilized combined with image processing technology to train a residual convolutional neural network integrated with an attention mechanism.With this core,detection soft⁃ware was developed for the detection of solid castings coupled with an image acquisition module to achieve non-contact measurement of casting surface roughness.The measured accuracy of the castings is 87.5%,with a single measure⁃ment of 0.086 s,which quickly and conveniently provides measurement results in actual production.

关 键 词：机器学习 表面粗糙度 图像处理 卷积神经网络 

分 类 号：TG247[金属学及工艺—铸造] TG242.1