An Efficient Approach to Remove Thick Cloud in VNIR Bands of Multi-Temporal Remote Sensing Images

Wenhui Du,Jinlong Fan,Zhihao Qin,Bilawal Abbasi,Fei Wang,Maofang Gao

doi:10.3390/rs11111284

Abstract

Cloud-free remote sensing images are required for many applications, such as land cover classification, land surface temperature retrieval and agricultural-drought monitoring. Cloud cover in remote sensing images can be pervasive, dynamic and often unavoidable. Current techniques of cloud removal for the VNIR (visible and near-infrared) bands still encounters the problem of pixel values estimated for the cloudy area incomparable and inconsistent with the cloud-free region in the target image. In this paper, we proposed an efficient approach to remove thick clouds and their shadows in VNIR bands using multi-temporal images with good maintenance of DN (digital number) value consistency. We constructed the spectral similarity between the target image and reference one for DN value estimation of the cloudy pixels. The information reconstruction was done with 10 neighboring cloud-free pair-pixels with the highest similarity over a small window centering the cloudy pixel between target and reference images. Four Landsat5 TM images around Nanjing city of Jiangsu Province in Eastern China were used to validate the approach over four representative surface patterns (mountain, plain, water and city) for diverse sizes of cloud cover. Comparison with the conventional approaches indicates high accuracy of the approach in cloud removal for the VNIR bands. The approach was applied to the Landsat8 OLI (Operational Land Imager) image on 29 April 2016 in Nanjing area using two reference images. Very good consistency was achieved in the resulted images, which confirms that the proposed approach could be served as an alternative for cloud removal in the VNIR bands using multi-temporal images.

Highlights

With development of earth observation, satellite remote sensing data have been extensively used for numerous studies [1,2], such as land surface temperature retrieval [3,4,5,6], agro-drought monitoring [7], soil moisture estimation [8], evapotranspiration modelling [9] and radiation flux estimation [10]
An efficient approach is proposed in this study to remove thick clouds along with their shadows
The basic idea is to reconstruct the cloud cover pixels of the target image with the DN values estimated with the linear regression modeling (LRM) method using several close cloud-free pixels centered the cloudy pixels of both target and multi-temporal reference images for each VNIR band

Summary

Introduction

With development of earth observation, satellite remote sensing data have been extensively used for numerous studies [1,2], such as land surface temperature retrieval [3,4,5,6], agro-drought monitoring [7], soil moisture estimation [8], evapotranspiration modelling [9] and radiation flux estimation [10]. In the application of satellite imagery, cloud-free images are strongly required in many studies, including land cover classification, time series analyses and estimation of Normalized Difference. Since the satellite imagery was acquired by the sensors of the platform in space, atmospheric effects are generally inevitable [14]. This is especially true in the tropical and mid-latitude regions [15]. Clouds and their shadows are significant noise sources for the use of the space-borne optical imagery, which can cause various problems in data analyses and applications, such as data fusion, land cover classification and land surface change monitoring. Zhu et al [20] improved the Fmask algorithm for Landsats 4–7 and developed a new version suitable for Landsat that takes advantage of the new cirrus band

Objectives

Methods

Findings

Discussion

Conclusion