Super-resolution generative adversarial network (SRGAN) enabled on-chip contact microscopy

This is the authors' abstract. We don't add key points for this paper.

We demonstrate a deep learning based contact imaging on a CMOS chip to achieve ∼1 μ m spatial resolution over a large field of view of ∼24 mm 2 . By using regular LED illumination, we acquire the single lower-resolution image of the objects placed approximate to the sensor with unit fringe magnification. For the raw contact-mode lens-free image, the pixel size of the sensor chip limits the spatial resolution. We apply a super-resolution generative adversarial networks, a type of deep learning based single-image super-resolution (SR) algorithm, to circumvent this limitation and effectively recover much higher resolution image of the objects, permitting sub-micron spatial resolution to be achieved across the entire sensor chip active area, which is also equivalent to the imaging field-of-view (24 mm 2 ) due to unit magnification. This SR contact imaging approach eliminates the need of either lens or multi-frame acquisition, being very powerful and cost-effective. We demonstrate the success of this approach by imaging the proliferation dynamics of large-scale cells and the Instantaneous behaviors of freely moving Caenorhabditis elegans directly on the chip.