Cloud Segmentation and Matching Using Deep Learning in All-Sky Images
DOI:
https://doi.org/10.52825/pv-symposium.v1i.1185Keywords:
All-Sky Imagers, Deep Learning, U-Net, Cloud SegmentationAbstract
In this paper, we focus on the segmentation of clouds in All Sky Images using a U-Net-based Deep Learning model and the subsequent recognition of the same cloud in different images. This research lays the foundation for the development of solar radiation forecasts with All-Sky Imagers. The implemented model initially extracts relevant features from the input image using convolutions, thereby reducing the resolution. In the subsequent step, the resolution is restored to its original level using transposed convolutions. Contours are then created from all segmented clouds. Using these contours as references, the same cloud is identified in images from different All-Sky Imagers through template and contour matching. We demonstrate that this segmentation approach yields good results on a small test dataset. Additionally, the recognition of clouds in images from different cameras show promising results, with 75 % of clouds being correctly matched.
Downloads
References
[1] M. Zehner, T. Weigl, M. Hartmann, S. Thaler, O. Schrank, B. Mayer, T. Betts, R. Gottschalg, K. Behrens, G. König-Langlo, B. Giesler, G. Becker und O. Mayer, „Energy Loss Due to Irradiance Enhancement,“ 2011.
[2] N. Stut, A. Boschert, F. Kaiser, M. Zehner, B. Mayer und O. Mayer, „Analyse von Einstrahlungsvolatilität und -überhöhungen in hochaufgelösten Datensätzen des DWD und MIM zur Untersuchung von Korrelationen zu meteorologischen Messdaten“.
[3] W. Xie, D. Liu, M. Yang, S. Chen, B. Wang, Z. Wang, Y. Xia, Y. Liu, Y. Wang und C. Zhang, „SegCloud: a novel cloud image segmentation model using a deep convolutional neural network for ground-based all-sky-view camera observation,“ Atmospheric Measurement Techniques, Bd. 13, p. 1953–1961, 2020. https://www.doi.org/10.5194/amt-13-1953-2020.
[4] D. Scaramuzza, A. Martinelli und R. Siegwart, „A Flexible Technique for Accurate Omnidirectional Camera Calibration and Structure from Motion,“ in Fourth IEEE International Conference on Computer Vision Systems (ICVS'06), New York, NY, USA, IEEE, 2006, pp. 45.
[5] D. Scaramuzza, A. Martinelli und R. Siegwart, „A Toolbox for Easily Calibrating Omnidirectional Cameras,“ in 2006 IEEE/RSJ International Conference on Intelligent Robots and Systems, Beijing, China, IEEE, 2006, pp. 5695-5701.
[6] M. Rufli, D. Scaramuzza und R. Siegwart, „Automatic detection of checkerboards on blurred and distorted images,“ 2008 IEEE/RSJ International Conference on Intelligent Robots and Systems, pp. 3121-3126, 2008.
[7] O. Ronneberger, P. Fischer und T. Brox, „U-Net: Convolutional Networks for Biomedical Image Segmentation,“ CoRR, Bd. abs/1505.04597, 2015.
[8] S. Suzuki und K. be, „Topological structural analysis of digitized binary images by border following,“ Computer Vision, Graphics, and Image Processing, Bd. 30, Nr. 1, pp. 32-46, 1985.
[9] M.-K. Hu, „Visual pattern recognition by moment invariants,“ IRE Transactions on Information Theory, Bd. 8, Nr. 2, pp. 179-187, 1962.
Published
How to Cite
Conference Proceedings Volume
Section
License
Copyright (c) 2024 Niklas Theis, Grit Behrens, Andreas Boschert, Mike Zehner
This work is licensed under a Creative Commons Attribution 4.0 International License.
Accepted 2024-08-07
Published 2024-08-19
