Multiplex optical sensors for reference structure tomography and compressive spectroscopy.

摘要

With the emergence of ubiquitous embedded digital processing and inexpensive focal planes, imaging and sensor systems that integrate optical and electronic processing have increasingly become attractive. These systems are regarded as "integrated computational imaging systems" because they use non-conventional optical elements to preprocess the field for digital analysis.;Computational imaging systems, also referred to as multiplex imagers, are particularly attractive as they achieve increased depth of field, multidimensional or multispectral imaging, improved data or computational efficiency and improved target recognition or tracking capabilities. Some of the capabilities of these computational systems are: multidimensional imaging, source analysis, data compression and analysis and sensor array data fusion.;In this thesis, a novel computational imaging system termed "Reference structure tomography (RST)" is proposed. RST makes use of a new form of volume optical element, a tomographic reference structure which modulates the visibility of the source space to improve the efficiency of object analysis. This thesis will describe RST and demonstrate three-dimensional imaging and source analysis with reference structures.;This thesis also proposes a new multiplex optical sensing system for compressive sampling in the physical layer. Compressive sampling enables signal reconstruction using fewer than one measurement per estimated signal value. A measurement efficient optical wavemeter and a compressive spectrometer will be described to demonstrate compressive sampling. The optical wavemeter efficiently measures the spectral densities of quasi-monochromatic and spectrally sparse sources. The compressive spectrometer measures the spectrum of a broad-band source with fewer measurements than the number of estimated spectral channels. The compressive spectrometer is designed to physically measure the Haar wavelet coefficients of the spectral source and the spectrum is estimated from these coefficients.