Narcissus Pattern Estimation for 2D-IRCCD Based FLIR

Abstract

This paper describes a method for estimating narcissus pattern of 2D-IRCCD based forward looking infrared camera (FLIR). Narcissus is a common optical noise in FUR caused by infrared emission from warm housing and cold surface reaching to the detector after reflected at lens surfaces. Since sharp edges in the narcissus pattern result in poor image quality, smooth narcissus edge is a significant design concept for FLIR. The narcissus pattern estimation method proposed in this paper hypothetically treats the cold surface as a negative infrared light source. Infrared intensity from this negative source creates narcissus at the detector surface. The narcissus pattern is obtained by normalizing the intensity variation of narcissus. A simulation program is developed using ray tracing technique for narcissus pattern calculation. Examples of both simulation and experimental results are presented.