The three-dimensional reconstruction of microstructures within transparent media based on in-situ incoherent tomographic imaging during laser processing

Internal microstructures within thick transparent media lack low-cost, high-precision in-situ detection methods. Affected by inherent limitations of defocus blur, widefield imaging systems struggle to accurately capture the internal microstructures of thick samples. Considering the exacerbation of spherical aberrations with increased imaging depth within the sample due to mismatch of refractive indices, and integrating apparatus parameters and image acquisition process, the three-dimensional point spread functions of a polychromatic point light source at various depths within the sample were calculate. By analyzing the axial relative intensity distribution of the calculated results, a dynamic programming three-dimensional deconvolution method was proposed for the restoration of tomographic images within thick transparent media. The restored original data from tomographic images is utilized for the three-dimensional reconstruction of internal fabricated microstructures. The accuracy of the in-situ detection method was validated by integrating a white light-emitting diode and an industrial camera on an ultrashort pulse laser-induced internal modification equipment of silicon carbide wafers. Compared to the detection results of scanning electron microscope, the microstructure reconstruction method achieves a reconstruction deviation of only 60 nm, at a cost less than 1/20th of the former. The in-situ detection method significantly enhances the accuracy of wide-field imaging systems in restoring original data of tomographic images for thick samples and can be widely used in the field of in-situ nondestructive testing of internal microstructures of thick transparent media.