We discuss methods of image reconstruction which lead to non-negative images: no negative values appear in the estimate or approximation to the unknown object. We describe methods which always lead to non-negative estimates of the object, characterize some regularized and non-regularized methods, and characterize those orthonormal systems whose partial series approximations to the object are always non- negative. This paper surveys the authors' recent research, adds some new results, and discusses related work.

Date Published: 19 July 1999
PDF: 2 pages
Proc. SPIE 3749, 18th Congress of the International Commission for Optics, (19 July 1999); doi: 10.1117/12.355059

Published in SPIE Proceedings Vol. 3749:
18th Congress of the International Commission for Optics
Alexander J. Glass; Joseph W. Goodman; Milton Chang; Arthur H. Guenther; Toshimitsu Asakura, Editor(s)