Research On The Three-dimensional Focal Stacks Imaging And Its Applications

Focal stack(FS)imaging method is a fast and effective three-dimensional(3D)imaging technique for large-size sparse samples,which has been widely used in visible light regime and recently used in synchrotron X-ray regime.In this method,a series of 2D microscopic images focused at different depths of the sample are collected by shifting the sample with fixed steps along the optical axis.Then the information inside the sample along the depth dimension is extracted and finally its 3D structure is reconstructed.Compared with nano-tomography(nano-CT),FS method doesn't need to rotate samples which avoids artifacts due to the large-angle rotation in CT experiments and the risk of collision with the upstream aperture during rotation.Furthermore,this method is free of refocusing and re-registration after each rotation required by nano-CT,which improves the experimental efficiency.FS method is easy-operation,efficient and suitable to 3D imaging of thick,wide and low-density sparse samples,meeting the user needs of fast 3D structure acquisition.The traditional FS reconstruction algorithm cannot reconstruct samples with complex structures,especially thick samples with multiple features along the same optical-axis pixel line(OAPL).Therefore,it needs to be further developed and improved.To overcome the disadvantages of traditional algorithms,we systematically studied the synchrotron-based 3D FS imaging methodology.Furthermore,a combination of the FS method and spectro-microscopy was developed,and applied to characterize the 3D spatial distributions of elements in various polymer samples.The main contributions in this dissertation are summarized as follows:(1)Improving the traditional FS algorithm and broadening the application scope of the method.In the traditional algorithm,the local variances of each pixel in the series of axially-stacked images is calculated,but only the maximal Rn along each OAPL is extracted as the focused sample feature.The algorithm is limited to single-layer or hollow samples.We proposed an improving scheme by thresholding the local variances with an iteratively calculated threshold Tn for each OAPL.Then all the positions and magnitudes of pixels/voxels with magnitudes higher than Tn were extracted and combined to form the 3D information of interest for the samples.(2)Performing simulations and online experiments to validate the new algorithm.The selection of the samples spans from simple to complex.The online experiments for the single layer Fe3O4 sample,the PVA air-filled microspheres and the polymer film embedded with Fe3O4 nanoparticles(NPs)were carried out on a scanning transmission X-ray microscope and the results reconstructed with the old and new algorithms were contrastively analyzed.It was found that the NPs in the sample will be partially lost with the traditional algorithm.(3)Developing a dual-energy FS method by combining the improved 3D FS method with the dual-energy contrast element mapping technique to characterize the 3D spatial distribution of the elements in samples.A dual-energy FS simulation experiment and an online experiment of porous polystyrene-doped fibers were performed to obtain the spatial distribution of iron acetylacetonate in the fibers,and analyze the oil absorption properties of the fibers from the perspective of 3D imaging.It provided a base for the subsequent water-oil separation experiments.The algorithm used in the dual-energy contrast analysis is the K-edge subtraction algorithm or thedual-energy ratio contrast algorithm.(4)Firstly applying the dual-energy FS method to the 3D imaging study of CNF/Fe3O4/TiO2(Cellulose nanofibrils,CNF)film samples.The 3D spatial distributions of Fe3O4 and TiO2 NPs were recognized clearly and quickly.The influence of relative proportions between Fe3O4 and TiO2 NPs on the mechanical,magnetic and light transmitting properties were investigated.It was found that the material properties of the composite films are different with different proportions of Fe3O4 and TiO2 NPs.The CNF/Fe3O4/TiO2 composite films with excellent magnetic,mechanical and anti-ultraviolet performances have many potential applications in the field of magneto-optical regimes.