Digital Confocal Microscopy And Research Of Its Image Restoration Algorithms

In biomedicine, a powerful tool for analyzing structure of cells, LCM (Laser Confocal Microscope), has been widely used; nevertheless it has a few disadvantages, especially its expensive cost, which make the limitation of its universal application. As a result, a novel technology, named DCM (Digital Donfocal Microscopy), can be used to obtain high-quality microscope images, which should firstly be recorded by focusing the conventional light microscope at different planes through the specimen along the light axis, then be deblurred by image restoration. Because the quality of images processed by DCM is even better than that of LCM's, DCM is generally and deeply studied rapidly.In this paper, the DCM is mainly described, especially an important part of which, image restoration, is discussed in detail. As follows work is achieved: first, the acquisition of series of specimen images is analyzed; then with the condition that assumes the PSF is known, the detected images are deblurred through methods of Nearest Neighbor, Wiener Filter, and Least Square respectively; at last, the degraded images are processed together with the unknown PSF is estimated by Wiener IBD, PBD, and NAS-RIF in turn, all of which get satisfying results.In this paper, it has been obtained a set of mature experiences and data on the parameters selected in PSF, Nearest Neighbor, Wiener filter, and Least Square; Furthermore, the methods determining the support constraints in IBD are developed by experiences. In PBD, the realization with the flow of PBD is presented, involving the modified optimization-searching algorithm that can improve its efficiency; Besides, the scheme extending the background of the image estimates is described in NAS-RIF. All of these will improve the efficiency of image restoration.In the end of this paper, it is concluded the disadvantages of DCM, promisingdirections and emphases in future research as well.