Computational Imaging

Computational Imaging

We are developing new approaches to imaging in which computational image recovery and novel optical systems work in unison to enable new imaging capabilities – systems that can record more information from a scene than can be obtained by conventional methods. For example, we have developed and experimentally demonstrated wavefront-­coded imaging systems at visible and thermal infrared wavelengths that can have figures of merit, such as depth of field or dimensions that are an order of magnitude better than can be achieved by conventional methods. The use of novel discontinuous lens designs is promising to enable gigapixel diffraction-limited imaging; or of simultaneous imaging and ranging in 3D. At the most computationally intensive extreme of computational imaging we research aperture synthetic imaging (at mm-­wavelengths) and compressive recording of images using single-­pixel detectors.

 

Above: theoretical demonstration of super resolution of an image using 25 low-resolution images.

Below: low cost multi-camera array for hyperspectral imaging.