Stray light and its reduction of focal plane arrays

Abstract

Existing studies have paid much attention to stray light of lenses. However, the ghost reflection inside FPA (Focal Plane Array) detector is almost ignored which is, in fact, also an unnegligible source of stray light in digital imaging systems. Ghosting between window surfaces and CCD photosensitive area is often a major source of stray light in FPA detectors, it may lead to image blur, color distortion, and contrast reduction. Besides, diffraction pattern can be observed from front-illuminated CCD as the incident beam can be diffracted by its polysilicon electrode gates at the surface. In this paper, we study the generation mechanism and the reduction approaches to such stray light. Both front-illuminated and back-illuminated CCDs (Charge Coupled Device) are investigated in our study. We build models to identify stray light paths and predict window ghost image characteristics. Furthermore, three methods, i.e., anti-reflective coating method, fluid-filled method and deconvolution method are presented to decrease the stray light. The first anti-reflective coating method can be effectively used in back-thinned CCDs to reduce the reflection and to maximize quantum efficiency. The second fluid-filled method attempts to reduce the effects of unwanted light contamination by simulating some characteristics of the human eyes. The use of liquid can reduce the fresnel reflectance of interface. In addition, the acquired images contaminated by FPA's stray light are post-processed with the deconvolution method. Effectiveness of our proposed methods is verified with experiments. It is shown that stray light of FPA can be efficiently reduced.