Session
Advanced Technologies 2
Location
Salt Palace Convention Center, Salt Lake City, UT
Abstract
For large-scale Earth observation constellations, such as Planet’s SkySat fleet, a major challenge is mitigating the effects of post-launch optical aberrations, which can render an asset unusable. This paper presents a novel approach to recover operational value from SkySat-C5, which remained excluded from production for 8 years due to significant, uncorrectable focus deficit at launch. Our method leverages stars to calibrate a reliable Point Spread Function (PSF) used in a deconvolution process that restored the camera’s images to a usable level of quality.
Because there is no on-orbit focus mechanism, the only path to restoration is post-processing deconvolution using a PSF that characterizes the telescope. Stars provide near-perfect optical point sources upon a black background, whereas terrestrial point sources are subject to noise and artifacts by atmospheric interference and background clutter. The result is a high quality PSF with excellent signal to background ratio (SBR) that characterizes the telescope’s optical characteristics.
The workflow to capture the point source required specialized operations to safely pitch the satellite away from the Earth to image the selected star, Betelgeuse. A tool called “SkyShot” was developed to manage the specialized concept of operations (ConOps) required to point the satellite and adjust the camera settings to successfully acquire the stellar target after ground and on-orbit experimentation.
The calibrated Point Spread Function (PSF) was utilized in a deconvolution process to restore image contrast and facilitate the efficient extraction of high-frequency information from the blurred captures. This process did not introduce new information but rather strategically processed existing information to maximize image quality. The resultant improvement in image processing yielded an absolute spatial resolution of approximately 1 meter in the native panchromatic band, thereby moving the satellite closer to achieving fleet-standard performance.
The result was a substantial enhancement in the satellite's imaging capacity, successfully bringing the asset back into production after years of on-orbit dormancy. We detail the end-to-end pipeline from satellite attitude control maneuvers and camera parameter optimization to the deconvolution algorithm and mitigation of sharpening artifacts. This success demonstrates a robust, repeatable methodology for maximizing image quality of degraded assets through stellar calibration.
Document Type
Event
Back From the Dead: Using Stellar PSF Calibration to Recover Operational Value From a Defocused SkySat Telescope
Salt Palace Convention Center, Salt Lake City, UT
For large-scale Earth observation constellations, such as Planet’s SkySat fleet, a major challenge is mitigating the effects of post-launch optical aberrations, which can render an asset unusable. This paper presents a novel approach to recover operational value from SkySat-C5, which remained excluded from production for 8 years due to significant, uncorrectable focus deficit at launch. Our method leverages stars to calibrate a reliable Point Spread Function (PSF) used in a deconvolution process that restored the camera’s images to a usable level of quality.
Because there is no on-orbit focus mechanism, the only path to restoration is post-processing deconvolution using a PSF that characterizes the telescope. Stars provide near-perfect optical point sources upon a black background, whereas terrestrial point sources are subject to noise and artifacts by atmospheric interference and background clutter. The result is a high quality PSF with excellent signal to background ratio (SBR) that characterizes the telescope’s optical characteristics.
The workflow to capture the point source required specialized operations to safely pitch the satellite away from the Earth to image the selected star, Betelgeuse. A tool called “SkyShot” was developed to manage the specialized concept of operations (ConOps) required to point the satellite and adjust the camera settings to successfully acquire the stellar target after ground and on-orbit experimentation.
The calibrated Point Spread Function (PSF) was utilized in a deconvolution process to restore image contrast and facilitate the efficient extraction of high-frequency information from the blurred captures. This process did not introduce new information but rather strategically processed existing information to maximize image quality. The resultant improvement in image processing yielded an absolute spatial resolution of approximately 1 meter in the native panchromatic band, thereby moving the satellite closer to achieving fleet-standard performance.
The result was a substantial enhancement in the satellite's imaging capacity, successfully bringing the asset back into production after years of on-orbit dormancy. We detail the end-to-end pipeline from satellite attitude control maneuvers and camera parameter optimization to the deconvolution algorithm and mitigation of sharpening artifacts. This success demonstrates a robust, repeatable methodology for maximizing image quality of degraded assets through stellar calibration.
