Hyperspectral Classification Onboard the Hypso-1 Cubesat.

Jonas G. Røysland, Dennis D. Langer, Simen Berg,Milica Orlandic,Joseph L. Garrett

Workshop on Hyperspectral Image and Signal Processing(2023)

引用 0|浏览0
暂无评分
摘要
A real-time hyperspectral classification pipeline, consisting of a support vector machine in a binary decision tree architecture, has been developed for the HYPSO-1 cubesat. On May 24, 2023, the pipeline classified its first image in orbit in less than ten seconds. In tests on the ground, prior to the deployment of the pipeline, both radiance and top-of-atmosphere reflectance data performed similarly well, and sparsity was shown to give a slight speed improvement.
更多
查看译文
关键词
Machine Learning,Remote Sensing,Support Vector Machines,Embedded Systems
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要