Resumen: Machine learning techniques are widely used for pixel-wise classification of hyperspectral images. These methods can achieve high accuracy, but most of them are computationally intensive models. This poses a problem for their implementation in low-power and embedded systems intended for on-board processing, in which energy consumption and model size are as important as accuracy. With a focus on embedded anci on-board systems (in which only the inference step is performed after an off-line training process), in this paper we provide a comprehensive overview of the inference properties of the most relevant techniques for hyperspectral image classification. For this purpose, we compare the size of the trained models and the operations required during the inference step (which are directly related to the hardware and energy requirements). Our goal is to search for appropriate trade-offs between on-board implementation (such as model size anci energy consumption) anci classification accuracy. Idioma: Inglés DOI: 10.3390/rs12030534 Año: 2020 Publicado en: Remote Sensing 12, 3 (2020), 534 [29 pp.] ISSN: 2072-4292 Factor impacto JCR: 4.848 (2020) Categ. JCR: GEOSCIENCES, MULTIDISCIPLINARY rank: 27 / 198 = 0.136 (2020) - Q1 - T1 Categ. JCR: IMAGING SCIENCE & PHOTOGRAPHIC TECHNOLOGY rank: 8 / 29 = 0.276 (2020) - Q2 - T1 Categ. JCR: REMOTE SENSING rank: 10 / 32 = 0.312 (2020) - Q2 - T1 Categ. JCR: ENVIRONMENTAL SCIENCES rank: 76 / 273 = 0.278 (2020) - Q2 - T1 Factor impacto SCIMAGO: 1.284 - Earth and Planetary Sciences (miscellaneous) (Q1)