Inversion of Rough Surface Parameters From SAR Images Using Simulation-Trained Convolutional Neural Networks

Song, Tao; Kuang, Lei; Han, Lingyan; Wang, Yuheng; Liu, Qing

doi:10.1109/lgrs.2018.2822821

Cited by 21 publications

(19 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Based on the above experimental results, we can summarize that the CNN approach can better learn the complex relationship between the Rrs and the OC-CCI Chla values than SVR, demonstrating the possibility that the CNN may be used for building an end-to-end approach for efficient Chla concentration estimation. Song et al 52 proposed a novel inversion method for rough surface parameters using CNN, which microwave images of rough surfaces are used for training. In Ref.…”

Section: Discussionmentioning

confidence: 99%

Global chlorophyll-a concentration estimation from moderate resolution imaging spectroradiometer using convolutional neural networks

Peng

et al. 2020

J. Appl. Rem. Sens.

View full text Add to dashboard Cite

The accurate estimation of global chlorophyll-a (Chla) concentration from the large remote sensing data in a timely manner is crucial for supporting various applications. Moderate resolution imaging spectroradiometer (MODIS) is one of the most widely used earth observation data sources, which has the characteristics of global coverage, high spectral resolution, and short revisit period. So the estimation of global Chla concentration from MODIS imagery in a fast and accurate manner is significant. Nevertheless, the estimation of Chla concentration from MODIS using traditional machine learning approaches is challenging due to their limited modeling capability to capture the complex relationship between MODIS spatial-spectral observations and the Chla concentration, and also their low computational efficiency to address large MODIS data in a timely manner. We, therefore, explore the potential of deep convolutional neural networks (CNNs) for Chla concentration estimation from MODIS imagery. The Ocean Color Climate Change Initiative (OC-CCI) Chla concentration image is used as ground truth because it is a well-recognized Chla concentration product that is produced by assimilating different satellite data through a complex data processing steps. A total of 12 monthly OC-CCI global Chla concentration maps and the associated MODIS images are used to investigate the CNN approach using a cross-validation approach. The classical machine learning approach, i.e., the supported vector regression (SVR), is used to compare with the proposed CNN approach. Comparing with the SVR, the CNN performs better with the mean log root-mean-square error and R 2 of being 0.129 and 0.901, respectively, indicating that using the MODIS images alone, the CNN approach can achieve results that is close to the OC-CCI Chla concentration images. These results demonstrate that CNNs may provide Chla concentration images that are reliable, stable and timely, and as such CNN constitutes a useful technique for operational Chla concentration estimation from large MODIS data. © The Authors. Published by SPIE under a Creative Commons Attribution 4.0 Unported License. Distribution or reproduction of this work in whole or in part requires full attribution of the original publication, including its DOI.

show abstract

Section: Discussionmentioning

confidence: 99%

Global chlorophyll-a concentration estimation from moderate resolution imaging spectroradiometer using convolutional neural networks

Peng

et al. 2020

J. Appl. Rem. Sens.

View full text Add to dashboard Cite

show abstract

“…As the typical deep neural networks, CNNs can make use of data in the form of spatially focused images. CNNs can even utilize complex microwave field data with rich information to extract features and realize recognition and imaging [34]- [40]. In fact, though facing huge noise and complex measured microwave field data, CNNs can still succeed in realizing the far-field imaging [34]- [40].…”

Section: B Proposed Cnn Architecturementioning

confidence: 99%

“…CNNs can even utilize complex microwave field data with rich information to extract features and realize recognition and imaging [34]- [40]. In fact, though facing huge noise and complex measured microwave field data, CNNs can still succeed in realizing the far-field imaging [34]- [40]. Hence, in the process of training, we convert simulated original propagating wave gain data as inputs to our deep CNNs, and use simulated subwavelength imaging as the output.…”

Section: B Proposed Cnn Architecturementioning

confidence: 99%

Applying Deep Learning Approach to the Far-Field Subwavelength Imaging Based on Near-Field Resonant Metalens at Microwave Frequencies

Yao

Jiang

2019

IEEE Access

View full text Add to dashboard Cite

In this paper, we utilize the deep learning approach for the subwavelength imaging in far-field, which is realized by the near-field resonant metalens at microwave frequencies. The resonating metalens consisting of split-ring resonators (SRRs) are equipped with the strong magnetic coupling ability and can convert evanescent waves into propagating waves using the localized resonant modes. The propagating waves in the far-field are utilized as the input of a trained deep convolutional neural network (CNN) to realize the imaging. The training data for establishing the deep CNN are obtained by the EM simulation tool. Besides, the white Gaussian noise is added into the training data to simulate the interference in the real application scenario. The proposed CNN composes of three pairs of convolutional and activation layers with one additional fully connected layer to realize the recognition, i.e., the imaging process. The feasibility of utilizing the trained deep CNN for imaging is validated by numerical benchmarks. Distinguished from the subwavelength imaging methods, the spatial response and Green's function need not be measured and evaluated in the proposed method. INDEX TERMS Convolutional neural network, resonant metalens, machine learning, subwavelength imaging.

show abstract

“…The method could still be applied to higher frequencies where the same criteria is still true, it would not be the case with much lower frequencies as the Fresnel zone would get wider. Directional antennas were used with a non negligible Rayleigh distance, which will have increased the diffraction loss compared to using omnidirectional antennas [3] though this effect is not of concern in this work. There are a number of physical parameters shown in Fig.…”

Section: Measurement Scenariomentioning

confidence: 99%

“…Two angles are defined as the elevation of the line of sight, θ LOS and the azimuth rotation of the body, θ BODY . The shield edge Fresnel diffraction parameters u and v defined in [3] based on the width of the body and its offset from the line of sight respectively will also affect the diffraction. To prove the concept in this paper, θ LOS and w are fixed at zero, while u is a constant value of 2.35 at the center frequency across the torso shown in Fig.…”

Section: Measurement Scenariomentioning

confidence: 99%

Machine-Learning-Based Approach for Diffraction Loss Variation Prediction by the Human Body

Khalily

Brown

Tafazolli

2019

Antennas Wirel. Propag. Lett.

View full text Add to dashboard Cite

This paper presents a machine learning (ML) based model to predict the diffraction loss around the human body. Practically, it is not reasonable to measure the diffraction loss changes for all possible body rotation angles, builds and line of sight (LoS) elevation angles. A diffraction loss variation prediction model based on a non-parametric learning technique called Gaussian process (GP) is introduced. Analysed results state that 86% correlation and normalised mean square error (NMSE) of 0.3 on the test data is achieved using only 40% of measured data. This allows a 60% reduction in required measurements in order to achieve a well-fitted ML loss prediction model. It also confirms the model generalizability for non-measured rotation angles.Index Terms-Diffraction loss, guassian process (GP), machine learning (ML), network planning tool.

show abstract

Inversion of Rough Surface Parameters From SAR Images Using Simulation-Trained Convolutional Neural Networks

Cited by 21 publications

References 17 publications

Global chlorophyll-a concentration estimation from moderate resolution imaging spectroradiometer using convolutional neural networks

Global chlorophyll-a concentration estimation from moderate resolution imaging spectroradiometer using convolutional neural networks

Applying Deep Learning Approach to the Far-Field Subwavelength Imaging Based on Near-Field Resonant Metalens at Microwave Frequencies

Machine-Learning-Based Approach for Diffraction Loss Variation Prediction by the Human Body

Contact Info

Product

Resources

About