Abstract

Remote sensing satellite images with high temporal and high spatial resolution play a critical role in earth science applications. However, it is difficult for a single satellite to obtain such images due to technical and cost constraints. Therefore, spatiotemporal image fusion based on deep learning has received extensive attention in recent years. This article proposes a multicooperative deep convolutional neural network (MCDNet) for spatiotemporal satellite image fusion. This method is a new multinetwork model in which multiple networks work together to reconstruct the predicted image. The multinetwork model consists of a super-resolution network, a difference reconstruction network, and a collaborative training network. First, the super-resolution network uses the combination of a novel multiscale mechanism and dilated convolutions to make full use of the spectral information of the coarse image and upgrade it to a transitional image that matches the fine image. The difference reconstruction network uses structural relevance to complete the reconstruction of the fine difference image. The collaborative training network extracts the hidden information from the fine image and uses the time relevance to restrict the training of the difference reconstruction network. Finally, the fine difference image and the known fine image are combined to complete the image fusion. The new compound loss function can help multinetwork models better complete cooperative training. Through experiments on two datasets and comparison with existing fusion algorithms, the subjective and objective results prove that MCDNet can effectively reconstruct higher-quality prediction images.

Highlights

  • IT is typically use dense time satellite data images with high spatial resolution in the dynamics of the earth system

  • To evaluate the performance of different spatiotemporal fusion algorithms in recent years and verify the effect of the fusion model proposed in this paper, we used two actual datasets for the experiments

  • In the first research dataset, the Landsat-5 Thematic Map collected images were used as fine images, and the moderate-resolution imaging spectroradiometer (MODIS) Terra MOD09GA Collection 5 images were used as coarse images

Read more

Summary

Introduction

IT is typically use dense time satellite data images with high spatial resolution in the dynamics of the earth system. Because of hardware technology and cost constraints, it is difficult for a single satellite to directly obtain dense time satellite data images with high spatial resolution [9], [10]. Some satellites have a shorter revisit period at the expense of the spatial resolution of the acquired data. Landsat can acquire hyperspectral images with a spatial resolution of 15 m or 30 m, but the revisit period is 16 days [12], [13]. Spatiotemporal image fusion is a technology used to synthesize dense time satellite data images with high spatial resolution from multi-source remote sensing data [14]-[16]. The two heterogeneous images are fused to derive fine images of dense time series

Objectives
Methods
Results
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call