Surface roughness classification using light scattering matrix and deep learning

作     者：SUN Hao TAN Wei RUAN YiXiao BAI Long XU JianFeng 

作者机构：State Key Laboratory of Intelligent Manufacturing Equipment and TechnologySchool of Mechanical Science and EngineeringHuazhong University of Science and TechnologyWuhan 430074China Wuhan Digital Design and Manufacturing Innovation Center Co.LtdWuhan 430074China 

出 版 物：《Science China(Technological Sciences)》 (中国科学（技术科学英文版）)

年 卷 期：2024年第67卷第2期

页      面：520-535页

核心收录：

学科分类：08[工学] 0804[工学-仪器科学与技术] 0805[工学-材料科学与工程（可授工学、理学学位）] 0803[工学-光学工程] 

基　　金：supported by the National Key R&D Program of China(Grant No.2020YFB1710400) the Key R&D Project of Hubei Province(Grant No.2023BAB067) 

主　　题：surface roughness FDTD simulation GHS theory deep learning light scattering matrix 

摘      要：High-quality optical components have been widely used in various applications;thus,extremely high beam quality is ***,surface roughness is a key indicator of the surface *** this study,the angular distribution of light scattering field intensity was obtained for surfaces having different roughness profiles based on the finite difference time domain(FDTD)method,and the results were compared with those obtained using the generalized Harvey-Shack(GHS)*** was shown that the FDTD approach can be used for an accurate simulation of the scattered field of a rough surface,and the superposition of results obtained from many surfaces that have the same roughness level was in good agreement with the result given by the analytic GHS model.A light scattering matrix(LSM)method was proposed based on the FDTD simulation results that could obtain rich surface roughness *** classification effect of LSM was compared with that of the single-incidence scattering distribution(SISD)based on a ResNet-50 deep learning *** classification accuracy of the model trained with the LSM dataset was obtained as 95.74%,which was 23.40%higher than that trained using the SISD ***,the effects of different noise types and filtering methods on the classification performance were analyzed,and the LSM was also shown to improve the robustness and generalizability of the trained surface roughness ***,the proposed LSM method has important implications for improving the data acquisition scheme of current light scattering measurement systems,and it also has the potential to be used for detection and characterization of surface defects of optical components.