A natural field of application for compressive sensing theory is imaging. Indeed, numerous compressive imaging (CI) systems and applications have been developed during the last few years. This work addresses the quantization effect in CI, which is fundamental for most CI architectures. In this paper, the implications of sensor quantization on universal CI are investigated theoretically and demonstrated with numerical experiments. It is shown that employing a CI framework may set severe requirements on the quantization depth of the optical sensor used. The quantization depth overhead requirement may be prohibitive in many optical imaging scenarios employing typical CI architectures. Practical solutions that significantly alleviate this requirement are suggested.
© 2013 Optical Society of America
Original Manuscript: December 21, 2012
Revised Manuscript: March 10, 2013
Manuscript Accepted: April 7, 2013
Published: May 8, 2013
Adrian Stern, Yigal Zeltzer, and Yair Rivenson, "Quantization error and dynamic range considerations for compressive imaging systems design," J. Opt. Soc. Am. A 30, 1069-1077 (2013)