Coded aperture imaging with a HURA coded aperture and a discrete pixel detector

Abstract

An investigation into the gamma ray imaging properties of a hexagonal uniformly redundant array (HURA) coded aperture and a detector consisting of discrete pixels constituted the major research effort associated with this thesis. Such a system offers distinct advantages for the development of advanced gamma ray astronomical telescopes in terms of the provision of high quality sky images in conjunction with an imager plane which has the capacity to reject background noise efficiently. Much of the research was performed as part of an ESA sponsored study into a prospective space astronomy mission (GRASP). The effort involved both computer simulations and a series of laboratory test images. A detailed analysis of the system point spread function (SPSF) of imaging planes which incorporate discrete pixel arrays is presented and the imaging quality quantified in terms of the signal to noise ratio (SNR). Computer simulations of weak point sources in the presence of detector background noise were also investigated and revealed that an inherent source of noise in the form of a coding error is present in the vast majority of observations, diminishing the quality of the deconvolved images. The coding errors could be completely removed for systems which employ hexagonal pixels, provided the mask element shadow area was correctly oriented and dimensioned with respect to the imager pixels. Theories developed during the study were evaluated by a series of experimental measurements with a s^Co gamma ray point source, an Anger camera detector, and a rotating HURA mask. These tests were complemented by computer simulations designed to reproduce, as close as possible, the experimental conditions. The 60° antisymmetry property of HURAs has also been employed to remove noise due to detector systematic effects present in the experimental images, and rendered a more realistic comparison of the laboratory tests with the computer simulations. Plateau removal and weighted deconvolution techniques were also investigated as methods for the reduction of the coding error noise associated with the gamma ray images. These techniques are shown to be effective, particularly when used together, but only on images in which coding errors are present. No improvement in image quality is offered by these techniques in the "perfect overlap" situation of the HPD.

More information

Identifiers

Catalogue record

Export record