Cloud detection in all-sky images via multi-scale neighborhood features and multiple supervised learning techniques

Cheng, Hsu-Yung; Lin, Chih Lung

doi:10.5194/amt-10-199-2017

Cited by 36 publications

(18 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Shi et al (2017) used a superpixel-based graph model (GM) to integrate multiple source information and proposed a new ground-based cloud detection algorithm to solve the problem that with a single information source, it is difficult to split the cloud from the clear sky. By analyzing components and different color spaces using partial least squares regression, Dev et al proposed a supervised segmentation framework to segment ground-based cloud pixels without any manually defined parameters (Dev et al, 2017a). Neto et al (2010) described a new segmentation algorithm using Bayesian inference and multidimensional Euclidean geometric distance to segment the cloud and sky patterns in image pixels on the RGB color space.…”

Section: Introductionmentioning

confidence: 99%

Diurnal and nocturnal cloud segmentation of all-sky imager (ASI) images using enhancement fully convolutional networks

et al. 2019

View full text Add to dashboard Cite

Abstract. Cloud segmentation plays a very important role in astronomical observatory site selection. At present, few researchers segment cloud in nocturnal all-sky imager (ASI) images. This paper proposes a new automatic cloud segmentation algorithm that utilizes the advantages of deep-learning fully convolutional networks (FCNs) to segment cloud pixels from diurnal and nocturnal ASI images; it is called the enhancement fully convolutional network (EFCN). Firstly, all the ASI images in the data set from the Key Laboratory of Optical Astronomy at the National Astronomical Observatories of Chinese Academy of Sciences (CAS) are converted from the red–green–blue (RGB) color space to hue saturation intensity (HSI) color space. Secondly, the I channel of the HSI color space is enhanced by histogram equalization. Thirdly, all the ASI images are converted from the HSI color space to RGB color space. Then after 100 000 iterative trainings based on the ASI images in the training set, the optimum associated parameters of the EFCN-8s model are obtained. Finally, we use the trained EFCN-8s to segment the cloud pixels of the ASI image in the test set. In the experiments our proposed EFCN-8s was compared with four other algorithms (OTSU, FCN-8s, EFCN-32s, and EFCN-16s) using four evaluation metrics. Experiments show that the EFCN-8s is much more accurate in cloud segmentation for diurnal and nocturnal ASI images than the other four algorithms.

show abstract

Section: Introductionmentioning

confidence: 99%

Diurnal and nocturnal cloud segmentation of all-sky imager (ASI) images using enhancement fully convolutional networks

et al. 2019

View full text Add to dashboard Cite

show abstract

“…Fully convolutional networks are a powerful visual deep learning algorithm for semantic segmentation (Shelhamer et al, 2017). Replacing the fully connected layers of the traditional convolutional neural networks (CNN) with the convolutional layers, the FCN reduces the network parameters, improves the segmentation speed and shows a good result on semantic segmentation through trained end-to-end, and pixels-to-pixels (Cheng and Lin, 2017). The basic components of FCN include convolutional layers, pooling layers, activation functions and deconvolutional layers, as shown in Fig.…”

Section: Efcnmentioning

confidence: 99%

Diurnal and nocturnal cloud segmentation of ASI images using enhancement fully convolutional networks

Shi¹,

Zhou²,

Qiu³

et al. 2019

Preprint

View full text Add to dashboard Cite

Abstract. Cloud segmentation plays a very important role in the astronomical observatory site selection. At present, few researchers segment cloud in the nocturnal All Sky Imager (ASI) images. This paper proposes a new automatic cloud segmentation algorithm which utilizes the advantages of deep learning fully convolutional networks (FCN) to segment cloud pixels from diurnal and nocturnal ASI images, named enhancement fully convolutional networks (EFCN). Firstly, all the ASI images in the data set from the Key Laboratory of Optical Astronomy at the National Astronomical Observatories of Chinese Academy of Sciences (CAS) are converted from the red-green-blue (RGB) color space to hue-saturation-intensity (HSI) color space. Secondly, the channel of the HSI color space is enhanced by the histogram equalization. Thirdly, all the ASI images are converted from the HSI color space to RGB color space. Then after 100000 times iterative training based on the ASI images in the training set, the optimum associated parameters of the EFCN-8s model are obtained. Finally, we use the trained EFCN-8s to segment the cloud pixels of the ASI image in the test set. In the experiments our proposed EFCN-8s was compared with other four algorithms (OTSU, FCN-8s, EFCN-32s, and EFCN-16s) using four evaluation metrics. Experiments show that the EFCN-8s is much more accurate in the cloud segmentation for diurnal and nocturnal ASI images than other four algorithms.

show abstract

“…Brown et al [24] developed an enhanced decision tree classifier for land cover recognition on multi-temporal remote sensing imagery and achieved an overall accuracy of 82% on ground truth data. Yung et al [25] used supervised learning technology with multi-resolution function to integrate Random Forest and Support Vector Machine to the remote sensing imagery. Xia [26] applied multi-scale Cascade Forest to remote sensing satellite imagery recognition.…”

Section: Introductionmentioning

confidence: 99%

Enhanced Multi-Dimensional and Multi-Grained Cascade Forest for Cloud/Snow Recognition Using Multispectral Satellite Remote Sensing Imagery

et al. 2021

View full text Add to dashboard Cite

Cloud/snow recognition is one application of satellite remote sensing imagery in natural disaster monitoring. Deep learning technology has contributed to the improvement of the performance of cloud/snow recognition. However, deep learning-based methods cannot well balance the performance and efficiency of cloud/snow recognition. In this paper, an augmented multi-dimensional and multi-grained Cascade Forest is proposed for cloud/snow recognition. The multi-dimensional deep forest structure with the representation learning ability allows it to capture the spatial and spectral information of cloud/snow satellite imagery accordingly equipped with good recognition efficiency. Besides, a simple augmentation Random Erasing method is introduced for enhancing the robustness of cloud/snow recognition. The experimental results on the HJ-1A/1B dataset show that the proposed method improves the performance of cloud/snow recognition by extracting spectral information from multi-spectral satellite imagery. In addition, based on the tree-based structure, the proposed method well balances the performance and efficiency of cloud/snow recognition, which can be considered as an alternative to the Neural Network for cloud/snow recognition.INDEX TERMS Cloud/snow recognition, multi-dimensional and multi-grained, Random Erasing, representation learning.

show abstract

Cloud detection in all-sky images via multi-scale neighborhood features and multiple supervised learning techniques

Cited by 36 publications

References 36 publications

Diurnal and nocturnal cloud segmentation of all-sky imager (ASI) images using enhancement fully convolutional networks

Diurnal and nocturnal cloud segmentation of all-sky imager (ASI) images using enhancement fully convolutional networks

Diurnal and nocturnal cloud segmentation of ASI images using enhancement fully convolutional networks

Enhanced Multi-Dimensional and Multi-Grained Cascade Forest for Cloud/Snow Recognition Using Multispectral Satellite Remote Sensing Imagery

Contact Info

Product

Resources

About