Virtual Microscopy with Extended Depth of Field

In this paper, we describe a virtual microscope system, based on JPEG 2000, which utilizes extended depth of field (EDF) imaging. Through a series of observer trials we show that EDF imaging improves both the local image quality of individual fields of view (FOV) and the accuracy with which the FOVs can be mosaiced (stitched) together. In addition, we estimate the required bit rate to adequately render a set of histology and cytology specimens at a quality suitable for on-line learning and collaboration. We show that, using JPEG 2000, we can efficiently represent high-quality, high-resolution colour images of microscopic specimens with less than 1 bit per pixel.