Method And Application Research On Computational Microscopy Based Onrecovering Wavefront

Optical equipments and instruments drive scientific research development,e.g.,biology,medical research,material science and mirco-machine technology much.However,in some industrial applications,such as field environmental monitoring,tele-medicine and agroforestry,there is an urgent need for microscopic imaging equipments with low cost,portablity and environmentally robustness.With the development of modern optical components,such as laser diodes(LDs),light-emitting diodes(LEDs),optical fibers,binary optics,CMOS/CCD image sensors and computational technology,the combination of inexpensive optical components and computational imaging algorithms might simply the optical setups and improve the imaging performance.In this way,replacing optical components by algorithms,makes the original complex optical microscopes more structurally simple and portable.In this thesis,we systematically explore how to use wavefront-recovering methods to realize the simplification,digitization,cheapness and portability of optical microscops,and improve the robustness of microscopic equipments.The innovations of this thesis are as follows:a.The scalar angular spectrum theorem,the multi-distance/multi-wavelength based Gerchberg-Saxton(G-S)iterative algorithms,the multi-distance/multi-wavelength based light transport-of-intensity-equations(TIE),the image autofocusing methods and their integrated algorithms are discussed.The integrated algorithms in this thesis,rely less on the fine hardwares compared to others,which would promote the computational microscopy to applications and practices.b.Inspired by digital imaging(digital microscopy)and dual-path digital holography,we compare several mathematic standrads for autofocusing algorithms.Based on an amount of experimental data,we think Gini-index-of-Gradients of a picture and Tamura-coefficients-of-Gradients of a picture are fit for single-path diffractive defocusing imaging,which is the main imaging mamner in this thesis.We also point out that autofocusing algorithms could be used to measure the defocusing distance and measre the illumination wavelength for partial coherent imaging.c.The effects of partial coherent light source,CMOS/CCD image sensors and defocusing distances on lens-less microscopic imaging resolution and autofocusing algorithms are systematically discussed.d.Further study about the applications of computational imaging in optical instruments,a multi-spectral lensless microscope and a portable multi-mode quantitative phase imaging microscope.Firstly,the multi-spectral lensless microscopic imaging device has the characteristics of a high resolution(～1 μm)and a large field-of-view(6.4 mm*4.8 mm).The wavelength autofocusing algorithm has a spectral resolution of～1 nm,which also provides a novel idea for wavelength calibrating.Based on a LED array,autofocusing algorithms and wavefront recovering algorithms,we build a portable multi-mode quantitative phase imaging microscope,which enables bright complex light-field microscopy imaging and dark field imaging.e.Using autofocusing algorithm,a portable and inexpensive method for measuring the bulk refractive index is proposed.Compared with the traditional Abbey refractometers,our method has a unlimited measurement range.The measurement range of the traditional Abbe refractometer is 1.3<n<1.8,while ours is 1<n<∝.