Session
Frank J. Redd Student Competition
Location
Utah State University, Logan, UT
Abstract
We present LOST: Open-source Star Tracker (LOST), a suite of star tracking software particularly suitable for small satellite missions with limited computing resources and low-cost cameras. LOST contains implementations of a number of previously-proposed star tracking algorithms and a flexible framework for running and evaluating these algorithms. Our evaluation finds that LOST's algorithms are simultaneously able to maintain a strong combination of accuracy, runtime, and memory usage. In scenarios representative of a low-cost star tracker, LOST correctly identifies over 95% of images, and importantly, performs the entire star tracking pipeline in less than 35 milliseconds on a Raspberry Pi while using less than 1 MiB of memory, backed by a < 350 KiB database. These results indicate that LOST could be ported to an embedded or radiation-hardened CPU and still perform well enough to meet the accuracy requirements of many missions.
LOST: An Open-Source Suite of Star Tracking Software
Utah State University, Logan, UT
We present LOST: Open-source Star Tracker (LOST), a suite of star tracking software particularly suitable for small satellite missions with limited computing resources and low-cost cameras. LOST contains implementations of a number of previously-proposed star tracking algorithms and a flexible framework for running and evaluating these algorithms. Our evaluation finds that LOST's algorithms are simultaneously able to maintain a strong combination of accuracy, runtime, and memory usage. In scenarios representative of a low-cost star tracker, LOST correctly identifies over 95% of images, and importantly, performs the entire star tracking pipeline in less than 35 milliseconds on a Raspberry Pi while using less than 1 MiB of memory, backed by a < 350 KiB database. These results indicate that LOST could be ported to an embedded or radiation-hardened CPU and still perform well enough to meet the accuracy requirements of many missions.
