Multispectral Three-Dimensional Imaging Using Chaotic Masks

Chaos in the form of random optical waves generated due to light scattering has always been considered undesirable in imaging systems. The above view changed upon the discovery of the capability to uniquely encode spatio-spectral information in chaos and decode selective spatio-spectral information, opening new pathways for imaging. In conventional lens-based imaging systems, multispectral imaging can be achieved using only a colour sensor that is sensitive to spectral changes. Colour sensors have a lower signal-to-noise ratio than monochrome sensors and are not desirable in light-sensitive applications. Using a chaotic mask and computational reconstruction method, it is possible to convert a monochrome sensor into a colour sensor. In this approach, the imaging process consists of two steps. In the first step, the point spread function (PSF) library for different depths and wavelengths is recorded using a chaotic mask. In the next step, a colour object is recorded using the same chaotic mask under the same experimental conditions. Computational techniques are used to process the recording of colour objects with the PSF library to reconstruct multispectral three-dimensional images of the object. In this chapter, the fundamental principles of imaging using chaotic masks, design of amplitude-and phase-type chaotic masks, fabrication methods for manufacturing chaotic masks, computational reconstruction methods and experimental results of regular and high-speed imaging are presented.