Three Dimensional Microscopy Image Restoration Based On Depth Variant Imaging Model

In the three-dimensional microscopy imaging, each one of the optical slices includes the in-focus and the out-of- focus structures, and in-focus structures appear sharp whereas out-of-focus structures appear blurry. In order to recover image, the laser scanning confocal microscopy (LSCM) or computational optical sectioning microscopy (COSM) can be used. COSM has the advantages of high signal-noise ratio, non-bleaching and low cost, in spite of the disadvantage of low speed.In some COSM image restoration algorithm, PSF was assumed to be space invariant to simplify the computation. But in the practical imaging, because the substance of specimen is various, the refractive index in different depth is different. As the refractive index mismatch between the immersion medium and the mounting medium of the specimen, the PSF is depth variant. In this case, the supposition of invariant PSF will lead to artificial restoration. Therefore, an expectation maximization (EM) algorithm based on the depth-variant imaging model is proposed in this paper. It can improve image resolution, especially in depth, but it would result in losing dim detail and enhancing very bright isolated spots. So a regularized EM algorithm is used to avoid such disadvantages and to recover the detail of image in the depth-variant imaging model in three-dimensional optical sectioning microscopy. At last, because the speed of computation in the EM algorithm and regularized EM algorithm is too slow, an expectation-maximization algorithm for image...