Research On The Key Techniques Of Opticalelement Defect Super-resolution Detection Basedon Structured Illumination

Surface defect detection of optical elements is an important step in the field of ultra-precision optical manufacturing.Among the surface defect detection techniques of optical elements,the microscopic imaging method is widely used for its high resolution.However,due to the limited resolution of the optical system,when the defect size reaches the sub-micron scale,the traditional microscopic imaging method will be difficult to realize the quantitative detection of optical element surface defects.In recent years,the rapid development of super-resolution microscopy technology brings a new opportunity to study the limit of diffraction.Due to its fast imaging speed,simple optical path structure and no special requirements for fluorescent materials,the obvious structured illumination microscope has gradually become one of the most popular super-resolution microscopy techniques.Aiming at the problem that it is difficult to accurately detect the micro-nano defects on the surface of optical elements by traditional microscopic imaging method,a scheme of super-resolution detection of micro-nano defects on the surface of optical elements by structured illumination is proposed in this paper.The imaging principle,image reconstruction algorithm and experimental optical path design are studied in depth.The main research contents of this paper are as follows:Firstly,according to the principle of structured illumination microscopy,the super-resolution imaging process of structured illumination is simulated by computer..The effects of various parameters of structured illumination(modulation,spatial frequency,initial phase)and noise on the quality of super-resolution imaging were simulated and analyzed.Simulation results show that the current mainstream initial phase estimation algorithm introduces strip-like artifacts into the reconstructed image when the modulation of fringe is lower than 0.06.Random noise is more distributed in the high frequency part of the sample after spectrum separation,Secondly,a phase estimation algorithm based on multi-image correlation processing in frequency domain is proposed to solve the problem of the increase of initial phase parameter estimation error in low-modulation systems.The results show that the algorithm can still keep the estimation error within 0.08 rad when the modulation is lower than 0.06,eliminate the artifacts in the reconstructed image under the modulation system,and improve the reconstruction quality of super-resolution image.To solve the problem that structured illumination imaging of transparent optical elements may lead to the reduction of signal to noise ratio(SNR),the preprocessing method of multi-image superposition with the same phase is adopted to suppress random noise.The results show that the peak signal to noise ratio of reconstructed images can be improved rapidly after the first ten superpositions;Thirdly,an interferometric structured illumination super-resolution microscopy system based on spatial light modulator is designed and constructed.The test results of 100 nm gold nanoparticles showed that the ultimate resolution of the system increased from 390.8nm to279.4nm.Imaging results of silicon substrate resolution plate verify the effectiveness of the MCF algorithm.Structured illumination imaging was carried out on the fused quartz glass plate with scratch defects spaced at 300 nm,which verified the system's ability to detect micro-nano defects in transparent optical elements.