Loading [MathJax]/extensions/MathMenu.js
A Multi-Temporal Convolutional Autoencoder Neural Network for Cloud Removal in Remote Sensing Images | IEEE Conference Publication | IEEE Xplore

A Multi-Temporal Convolutional Autoencoder Neural Network for Cloud Removal in Remote Sensing Images


Abstract:

The uncontrollable weather conditions can cause a serious problem to remote sensing imaginary. One of the weather conditions is a resulting from cloud contamination. As a...Show More

Abstract:

The uncontrollable weather conditions can cause a serious problem to remote sensing imaginary. One of the weather conditions is a resulting from cloud contamination. As a result, this paper proposed the use of the convolutional autoencoder neural networks to remove clouds from cloud-contaminated images by training on a multi-temporal remote sensing dataset. Here, the observations from different spectral bands are assumed to be independent since their spectral responses are usually nonoverlapped. From this assumption, each convolutional autoencoder neural networks are trained with the observation from only one spectral band. In our method, we have three convolutional autoencoder neural networks for red, green and blue spectral bands. The experiments were conducted on both synthesis and real dataset derived from the actual LANDSAT 8 images from the central part of Thailand where our algorithm has shown to have a superb performance.
Date of Conference: 18-21 July 2018
Date Added to IEEE Xplore: 20 January 2019
ISBN Information:
Conference Location: Chiang Rai, Thailand

I. Introduction

In the last decades, satellite images have become one of the most powerful tools for monitoring earth ecosystem. Their utilization were found in various fields such as land cover observation, hazard monitoring, crop forecasting, urbanization and military operation. However, an appearance of clouds on satellite images can severely degrade the quality, and limit the use of satellite images. Thus, a cloud removal is a significant task that must be addressed in order to recover a large number of unused satellite images suffered by cloud contamination. However, cloud removal is a challenging task since the land cover materials underneath the clouds are often unknown or changed over time. Furthermore, even the images that acquired from the same sensors, the reflectance of the surface can be different due to details changing on the surface such as seasonal plantation, urban expansion and human activities which can be very difficult to predict.

References

References is not available for this document.