Paper Abstract and Keywords |
Presentation |
2022-12-16 10:45
Machine learning for global detection of photovoltaic panel installation using Landsat-8 Imagery Ryo Ito, Ryu Sugimoto, Chiaki Tsutsumi, Ryosuke Nakamura (AIST) SANE2022-77 |
Abstract |
(in Japanese) |
(See Japanese page) |
(in English) |
Nowadays, free-and-open medium-resolution satellite imagery can be utilized for sustainable monitoring of global changes. In this study, we investigate a method to detect and identify solar panels installed during two periods from Landsat8 imagery. In recent years, AI such as deep learning has been considered one of the monitoring methods. However, medium-resolution satellite imagery has some unique characteristics that become advantages and disadvantages compared to aerial imagery when applying AI. In the case of Landsat-8 imagery, it has multiple bands outside the visible range, but the 30m spatial resolution sometimes can obscure the shapes of features. In this study, we introduce several approaches based on these characteristics. |
Keyword |
(in Japanese) |
(See Japanese page) |
(in English) |
image identification / satellite image / Deep Learning / CNN / Landsat8 / solar panel / global changes / land cover change |
Reference Info. |
IEICE Tech. Rep., vol. 122, no. 312, SANE2022-77, pp. 74-76, Dec. 2022. |
Paper # |
SANE2022-77 |
Date of Issue |
2022-12-08 (SANE) |
ISSN |
Online edition: ISSN 2432-6380 |
Copyright and reproduction |
All rights are reserved and no part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Notwithstanding, instructors are permitted to photocopy isolated articles for noncommercial classroom use without fee. (License No.: 10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) |
Download PDF |
SANE2022-77 |
|