Capture and processing of 3d microscopic images through multi-perspective technology

Résumé

During the last decade new technologies for the acquisition of 3D images have shown an impressive growth. One of these techniques that is worth mentioning, due to its capability of capturing the 3D information in a single shot, is known under different names such as Integral Imaging, Plenoptic Imaging and Lightfield Imaging. Since their invention at the beginning of the 20th Century but mainly after their rebirth in the ‘90s, lightfield imaging systems have gathered the attention of a vast community of researchers thanks to their promising capabilities of capturing 3D structure of incoherently illuminated scenes with just a single shot. These results are achievable thanks to the capability of these systems to capture not only the spatial information of the light rays emitted by the scene, but also its angular information. This has opened new research paths towards the design of improved systems, new dedicated algorithms, and a great amount of new applications, that can vary from phone cameras for bokeh effect, till cinema production for after effects. Lately, Integral-Imaging systems have shown very promising capabilities of capturing the 3D structure of microscopic samples. Nevertheless, there are some technical limitations inherent to this technology that needs to be taken into account. In this Thesis we will analyse the theoretical principles of lightfield microscopy with particular focus to its bottleneck limitations with the scope of implementing new design solutions in order to overcome those problems. The aim of this work is to provide an optimal design for 3D-integral microscopy with extended depth of field and enhanced lateral resolution. The principal focus of this Thesis has been to contribute making a step forward to the lightfield microscopy technique, in both directions: optical optimization of the capturing system and development of new algorithms for the reconstruction of the 3D sample.