CloudSatNet-1: FPGA-Based Hardware-Accelerated Quantized CNN for Satellite On-Board Cloud Coverage Classification

Simple item page
dc.contributor.authorPitoňák, Radoslavcs
dc.contributor.authorMucha, Jáncs
dc.contributor.authorDobiš, Lukášcs
dc.contributor.authorJavorka, Martincs
dc.contributor.authorMarušin, Marekcs
dc.coverage.issue13cs
dc.coverage.volume14cs
dc.date.issued2022-07-02cs
dc.description.abstractCubeSats, the nanosatellites and microsatellites with a wet mass up to 60 kg, accompanied by the cost decrease of accessing the space, amplified the rapid development of the Earth Observation industry. Acquired image data serve as an essential source of information in various disciplines like environmental protection, geosciences, or the military. As the quantity of remote sensing data grows, the bandwidth resources for the data transmission (downlink) are exhausted. Therefore, new techniques that reduce the downlink utilization of the satellites must be investigated and developed. For that reason, we are presenting CloudSatNet-1: an FPGA-based hardware-accelerated quantized convolutional neural network (CNN) for satellite on-board cloud coverage classification. We aim to explore the effects of the quantization process on the proposed CNN architecture. Additionally, the performance of cloud coverage classification by biomes diversity is investigated, and the hardware architecture design space is explored to identify the optimal FPGA resource utilization. Results of this study showed that the weights and activations quantization adds a minor effect on the model performance. Nevertheless, the memory footprint reduction allows the model deployment on low-cost FPGA Xilinx Zynq-7020. Using the RGB bands only, up to 90% of accuracy was achieved, and when omitting the tiles with snow and ice, the performance increased up to 94.4% of accuracy with a low false-positive rate of 2.23% for the 4-bit width model. With the maximum parallelization settings, the hardware accelerator achieved 15 FPS with 2.5 W of average power consumption (0.2 W increase over the idle state).en
dc.formattextcs
dc.format.extent1-21cs
dc.format.mimetypeapplication/pdfcs
dc.identifier.citationRemote Sensing. 2022, vol. 14, issue 13, p. 1-21.en
dc.identifier.doi10.3390/rs14133180cs
dc.identifier.issn2072-4292cs
dc.identifier.orcid0000-0001-5126-440Xcs
dc.identifier.other178496cs
dc.identifier.researcheridT-9091-2019cs
dc.identifier.scopus57201029686cs
dc.identifier.urihttp://hdl.handle.net/11012/208170
dc.language.isoencs
dc.publisherMDPIcs
dc.relation.ispartofRemote Sensingcs
dc.relation.urihttps://www.mdpi.com/2072-4292/14/13/3180cs
dc.rightsCreative Commons Attribution 4.0 Internationalcs
dc.rights.accessopenAccesscs
dc.rights.sherpahttp://www.sherpa.ac.uk/romeo/issn/2072-4292/cs
dc.rights.urihttp://creativecommons.org/licenses/by/4.0/cs
dc.subjectCNNen
dc.subjectFPGAen
dc.subjecthardware acceleratorsen
dc.subjectimage processingen
dc.subjecton-board processingen
dc.subjectquantizationen
dc.titleCloudSatNet-1: FPGA-Based Hardware-Accelerated Quantized CNN for Satellite On-Board Cloud Coverage Classificationen
dc.type.driverarticleen
dc.type.statusPeer-revieweden
dc.type.versionpublishedVersionen
sync.item.dbidVAV-178496en
sync.item.dbtypeVAVen
sync.item.insts2025.02.03 15:42:27en
sync.item.modts2025.01.17 16:44:56en
thesis.grantorVysoké učení technické v Brně. Fakulta elektrotechniky a komunikačních technologií. Ústav telekomunikacícs
Files
Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
remotesensing1403180v2.pdf
Size:
24.63 MB
Format:
Adobe Portable Document Format
Description:
remotesensing1403180v2.pdf
Download
Collections
Ústav telekomunikací