Multilevel Cloud Detection for High-Resolution Remote Sensing Imagery Using Multiple Convolutional Neural Networks

Yang Chen,Xiucheng Yang,Wei Li,Muhammad Bilal,Rongshuang Fan,Jingxue Wang

doi:10.3390/ijgi7050181

Abstract

In high-resolution image data, multilevel cloud detection is a key task for remote sensing data processing. Generally, it is difficult to obtain high accuracy for multilevel cloud detection when using satellite imagery which only contains visible and near-infrared spectral bands. So, multilevel cloud detection for high-resolution remote sensing imagery is challenging. In this paper, a new multilevel cloud detection technique is proposed based on the multiple convolutional neural networks for high-resolution remote sensing imagery. In order to avoid input the entire image into the network for cloud detection, the adaptive simple linear iterative clustering (A-SCLI) algorithm was applied to the segmentation of the satellite image to obtain good-quality superpixels. After that, a new multiple convolutional neural networks (MCNNs) architecture is designed to extract multiscale features from each superpixel, and the superpixels are marked as thin cloud, thick cloud, cloud shadow, and non-cloud. The results suggest that the proposed method can detect multilevel clouds and obtain a high accuracy for high-resolution remote sensing imagery.

Highlights

With the advancement in remote sensing technology, high-resolution satellite imagery is widely used in various fields such as resource surveying, environmental monitoring, and geographical mapping [1]
The overall accuracy and kappa of the proposed approach was more than 95% (Table 3), and the overall edge accuracy was more than 97.37%, indicating that self-adaptive pooling and superpixel combinations are effective in multilevel cloud detection
This paper presents a cloud detection for high-resolution remote sensing imagery using and improved convolutional neural network model

Summary

Introduction

With the advancement in remote sensing technology, high-resolution satellite imagery is widely used in various fields such as resource surveying, environmental monitoring, and geographical mapping [1]. According to the International Satellite Cloud Climatology Project-Flux Data (ISCCP-FD), the global annual mean cloud cover is approximately 66% [2]. Cloud often appears and covers objects on the surface in high-resolution remote sensing images, which leads to missing information and spectral distortion, and can affect the processing of remote sensing imagery [3]. Cloud detection in high-resolution satellite imagery is of great significance. The threshold-based methods are practical and fast in calculation, so are widely used in practical applications. Threshold-based methods, including the International Satellite Cloud Climatology

Methods

Findings

Discussion

Conclusion