Direct Image Classification from Fourier Ptychographic Microscopy Measurements without Reconstruction

The computational imaging technique of Fourier Ptychographic Microscopy (FPM) enables high-resolution imaging with a wide field of view and can serve as an extremely valuable tool, e.g. in the classification of cells in medical applications. However, reconstructing a high-resolution image from tens or even hundreds of measurements is computationally expensive, particularly for a wide field of view. Therefore, in this paper, we investigate the idea of classifying the image content in the FPM measurements directly without performing a reconstruction step first. We show that Convolutional Neural Networks (CNN) can extract meaningful information from measurement sequences, significantly outperforming the classification on a single band-limited image (up to 12 %) while being significantly more efficient than a reconstruction of a high-resolution image. Furthermore, we demonstrate that a learned multiplexing of several raw measurements allows maintaining the classification accuracy while reducing the amount of data (and consequently also the acquisition time) significantly.

Light-field Microscopy with a Consumer Light-field Camera

We explore the use of inexpensive consumer light- field camera technology for the purpose of light-field mi- croscopy. Our experiments are based on the Lytro (first gen- eration) camera. Unfortunately, the optical systems of the Lytro and those of microscopes are not compatible, lead- ing to a loss of light-field information due to angular and spatial vignetting when directly recording microscopic pic- tures. We therefore consider an adaptation of the Lytro op- tical system. We demonstrate that using the Lytro directly as an oc- ular replacement, leads to unacceptable spatial vignetting. However, we also found a setting that allows the use of the Lytro camera in a virtual imaging mode which prevents the information loss to a large extent. We analyze the new vir- tual imaging mode and use it in two different setups for im- plementing light-field microscopy using a Lytro camera. As a practical result, we show that the camera can be used for low magnification work, as e.g. common in quality control, surface characterization, etc. We achieve a maximum spa- tial resolution of about 6.25{\mu}m, albeit at a limited SNR for the side views.