Ata uhi

Hyper Spectral Camera ANalyzer (HyperSCAN)

I tiakina i:
I whakaputaina i:Remote Sensing vol. 17, no. 5 (2025), p. 842
Kaituhi matua: Wen-Qian, Chang
Ētahi atu kaituhi: Hou, Hsun-Ya, Pei-Yuan, Li, Shen, Michael W, Cheng-Ling, Kuo, Tang-Huang, Lin, Chang, Loren C, Chi-Kuang, Chao, Liu, Jann-Yenq
I whakaputaina:
MDPI AG
Ngā marau:
Optical design
Optical components
Environmental monitoring
Students
Data processing
Artificial neural networks
Cameras
Disaster management
Spatial discrimination
Colleges & universities
Data integration
Research & development > R&D
Emergency preparedness
Optics
Deep learning
Cost effectiveness
Resource exploration
Climate studies
Earth
Spatial data
Spatial resolution
Sensors
Design
Modular design
Optical data processing
Mineral exploration
Image quality
Cubesat
Education
Neural networks
Hyperspectral imaging
Urunga tuihono:Citation/Abstract
Full Text + Graphics
Full Text - PDF
Ngā Tūtohu: Tāpirihia he Tūtohu
Kāore He Tūtohu, Me noho koe te mea tuatahi ki te tūtohu i tēnei pūkete!
Whakarāpopotonga:HyperSCAN (Hyper Spectral Camera ANalyzer) is a hyperspectral imager which monitors the Earth’s environment and also an educational platform to integrate college students’ ideas and skills in optical design and data processing. The advantages of HyperSCAN are that it is designed for modular design, is compact and lightweight, and low-cost using commercial off-the-shelf (COTS) optical components. The modular design allows for flexible and rapid development, as well as validation within college lab environments. To optimize space utilization and reduce the optical path, HyperSCAN’s optical system incorporates a folding mirror, making it ideal for the constrained environment of a CubeSat. The use of COTS components significantly lowers pre-development costs and minimizes associated risks. The compact size and cost-effectiveness of CubeSats, combined with the advanced capabilities of hyperspectral imagers, make them a powerful tool for a broad range of applications, such as environmental monitoring of Earth, disaster management, mineral and resource exploration, atmospheric and climate studies, and coastal and marine research. We conducted a spatial-resolution-boost experiment using HyperSCAN data and various hyperspectral datasets including Urban, Pavia University, Pavia Centre, Botswana, and Indian Pines. After testing various data-fusion deep learning models, the best image quality of these methods is a two-branches convolutional neural network (TBCNN), where TBCNN retrieves spatial and spectral features in parallel and reconstructs the higher-spatial-resolution data. With the aid of higher-spatial-resolution multispectral data, we can boost the spatial resolution of HyperSCAN data.
ISSN:2072-4292
DOI:10.3390/rs17050842
Puna:Advanced Technologies & Aerospace Database