Correcting spacecraft jitter in HiRISE images

• Published in: Planetary Remote Sensing and Mapping pp. 91 - 106
• Main Authors: Sutton, S. S., Boyd, A. K., Kirk, R. L., Cook, D., Backer, J. W., Fennema, A., Heyd, R., McEwen, A. S., Mirchandani, S. D.
• Format: Book Chapter
• Language: English
• Published: CRC Press 2019
• Edition: 1
• Subjects: Complementary Metal Oxide Semiconductor Detectors; Pixel Offsets; LRO; Cross-track Direction; High Resolution DEM; MRO; Jitter Measurement; Geometric Distortions; RFM; Boxcar Filter; Jitter Frequencies; Imu Data; DTM Quality; HiRISE; Nominal Surface; Color Registration; DTM Production; Jitter Amplitude; Active Pixel Sensors; Binary Large Object; Pushbroom Imaging; Jitter Plots; Image Strips; Line Datasets; Line Scan Camera
• ISBN: 9781138584150; 1138584150
• DOI: 10.1201/9780429505997-8

Mechanical oscillations or vibrations on spacecraft, also called pointing jitter, cause geometric distortions and/or smear in high-resolution digital images acquired from orbit. Geometric distortion is especially a problem with pushbroom sensors, such as the High Resolution Imaging Science Experiment (HiRISE) instrument on-board the Mars Reconnaissance Orbiter (MRO). Geometric distortions occur at a range of frequencies that may not be obvious in the image products, but can cause problems with stereo image correlation in the production of digital elevation models, and in measuring surface changes in time series with orthorectified images. The HiRISE focal plane comprises a staggered array of fourteen charge-coupled devices (CCDs) with pixel instantaneous field of view (IFOV) of 1 microradian. The high spatial resolution of HiRISE makes it both sensitive to, and an excellent recorder of jitter. We present an algorithm using Fourier analysis to resolve the jitter function for a HiRISE image that is then used to update instrument pointing information to remove geometric distortions from the image. Implementation of the jitter analysis and image correction is performed on selected HiRISE images made available to the public. Results show marked reduction of geometric distortions. This work has applications to similar cameras operating now (such as the Lunar Reconnaissance Orbiter Camera Narrow Angle Camera (LROC NAC) on-board the Lunar Reconnaissance Orbiter) and to the design of future instruments (such as the Europa Imaging System, planned for the Europa Clipper mission). The sample and line jitter functions are described separately. This text file is read into the subsequent step that uses the derived function to transform the images. The jitter corrections are applied and used to update the SPICE before finally projecting all the redundant (RED) image strips with noproj. This transformation is done by projecting pixels from the real camera down to a nominal surface and back up into the ideal and smoothly moving camera. Jitter amplitude is affected by binning, because the binned data are rescaled to match the bin mode of the RED image strips during the color registration process. Future instruments could use complementary metal-oxide semiconductor detectors that allow for customizing image line spacing to optimize the ability to resolve non-aliased jitter frequencies. Use of jitter-corrected HiRISE images shows marked improvements in Digital Terrain Model quality, specifically in terrain models that were previously rendered unacceptable by jitter-induced distortions.