The computational array camera

Computational   cameras have become ubiquitous in the research community (though   not yet in the consumer space), as various architectures are being considered for creating output features that are not possible, or very difficult to obtain with "traditional" cameras. Computational   cameras reflect our desire to capture more information about the world around us through imaging (whether in the visible spectrum or not), and to exploit that information in applications that enhance our interaction with the imaged scene.     
 
In this context, we will introduce the topic by discussing the definition and a taxonomy of computational cameras. Some examples of trade-offs that computational cameras make to achieve their unique features,   will be discussed. The unifying view that computational cameras can be seen as imaging codecs will be advanced, with the "encoder"   (e.g., modified optics) acting in well-defined ways on the imaged scene signal, and the attendant tightly-coupled digital processing being the "decoder" in charge of producing the output features of interest. This concept will be exemplified using various kinds of computational cameras. In this general framework, we will then follow with a review of a computational array camera developed by Pelican Imaging, Inc. Some of the capabilities of this array camera will be illustrated.

Speaker: Dan Lelescu, Pelican Imaging