Rochester Institute of Technology
Sub-pixel event repositioning algorithms for Chandra X-ray astronomical imaging
Abstract
dc:description.abstractChandra X-ray Observatory (CXO) with the Advanced CCD Imaging Spectrometer (ACIS) offers unprecedented spatial resolution for X-ray astronomical imaging. As focal plane arrays, ACIS CCDs were fabricated to satisfy CXO's requirements in spatial and spectral specifications, and operate in photon-counting mode to record any photoelectric interactions into event lists with energy information and charge split morphology (grade). The charge cloud generated by a single X-ray photon has a relatively small size compared to ACIS pixel, and the shape is approximated by an axial-symmetric Gaussian with full width half maximum (FWHM) about 2 to 4 microns. This fact indicates that the impact positions of photons that generate split events are near the pixel boundaries, instead of the pixel centers. Considering ACIS CCDs have pixel size of 24 microns, subpixel event repositioning (SER) algorithms designed to refine the positions of split events should significantly improve the spatial resolution of Chandra/ACIS imaging. SER methods have been modified, from original corner-split events only model, to static SER (including 2-pixel and single pixel events), energy-dependent SER, and charge-split dependent SER, for both backside-illuminated (BI) and frontside-illuminated (FI) CCDs. Both simulated and CXO-observation data demonstrate the improvement for various SER methods. Chandra/ACIS data obtained for the Orion Nebula Cluster (ONC) was used to evaluate the SER algorithms, by reconstructing point-like sources in ONC and measuring their FWHM before and after applying SER methods. The improvement of FWHM for simulated and ONC sources was analyzed, so as to establish the degree of image improvement achieved by, as well as limitations on the success of, subpixel event repositioning algorithms. BI and FI CCDs exhibit different performance and, overall, BI applications benefit more from angular resolution improvement after applying SER techniques. The best performance after applying SER techniques can be as much as 62%, i.e., very close to theoretically available improvement, depending on applied SER method, source spectrum, off-axis angle, and employed CCD type.
Degree
thesis:*- Name thesis:degree_name
- Imaging Science (Ph.D.)
- Level thesis:degree_level
- Dissertation
- Discipline thesis:degree_discipline
- Chester F. Carlson Center for Imaging Science (COS)
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Li, Jingqiang
- Contributors dc:contributor
-
- Joel Kastner
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.rit.edu/theses/7676