Simultaneous Estimation of Land Surface and Atmospheric Parameters From Thermal Hyperspectral Data Using a LSTM–CNN Combined Deep Neural Network

Ye, Xin; Ren, Huazhong; Nie, Jing; Jian, Hui; Jiang, Chenchen; Zhu, Jinshun; Wang, Fan; Ye, Qian; Liang, Yanzhen

doi:10.1109/lgrs.2021.3104501

Cited by 10 publications

(5 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To improve accuracy further, multiple LST estimation models were considered to be trained in different regions and various TOA BT ranges to describe the relationship between the TOA BT and LST more accurately and reduce the interference caused by spatial heterogeneity. Furthermore, an end-to-end deep learning model was considered to automatically extract the deep features between the TOA BT of different channels using neural networks [21,26] instead of manual feature engineering.…”

Section: Discussionmentioning

confidence: 99%

“…Benefitting from the ability to capture nonlinear correlations, machine learning methods, which can provide a direct solution by establishing a link between the input feature set to the output results, have been applied for parameter retrieval from TIR remote sensing data [20], including LST estimation from multiple-channel TIR data [21,22], LST residual optimization from SW algorithm results [23], and the simultaneous retrieval of land surface and atmospheric parameters from TIR hyperspectral data [24][25][26]. Ensemble learning methods, which can learn multiple hypotheses to solve a problem together by combining various weak learning models, including bagging, boosting, and stacking techniques, have demonstrated the ability to represent complex physical processes well and have been applied to remote sensing applications with promising results, such as mapping natural hazards, predicting crop yields, and spatial downscaling [20,27,28], and they can improve accuracy and reduce overfitting by learning multiple hypotheses from training datasets [29].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Land Surface Temperature Estimation from Landsat-9 Thermal Infrared Data Using Ensemble Learning Method Considering the Physical Radiance Transfer Process

Ye,

Liu,

Hui

et al. 2023

Land

View full text Add to dashboard Cite

Accurately estimating land surface temperature (LST) is a critical concern in thermal infrared (TIR) remote sensing. According to the thermal radiance transfer equation, the observed data in each channel are coupled with both emissivity and atmospheric parameters in addition to the LST. To solve this ill-posed problem, classical algorithms often require the input of external parameters such as land surface emissivity and atmospheric profiles, which are often difficult to obtain accurately and timely, and this may introduce additional errors and limit the applicability of the LST retrieval algorithms. To reduce the dependence on external parameters, this paper proposes a new algorithm to directly estimate the LST from the top-of-atmosphere brightness temperature in Landsat-9 two-channel TIR data (channels 10 and 11) without external parameters. The proposed algorithm takes full advantage of the adeptness of the ensemble learning method to solve nonlinear problems. It considers the physical radiance transfer process and adds the leaving-ground bright temperature and atmospheric water vapor index to the input feature set. The experimental results show that the new algorithm achieves accurate LST estimation results compared with the ground-measured LST and is consistent with the Landsat-9 LST product. In subsequent work, further studies will be undertaken on developing end-to-end deep learning models, mining more in-depth features between TIR channels, and reducing the effect of spatial heterogeneity on accuracy validation.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Land Surface Temperature Estimation from Landsat-9 Thermal Infrared Data Using Ensemble Learning Method Considering the Physical Radiance Transfer Process

Ye,

Liu,

Hui

et al. 2023

Land

View full text Add to dashboard Cite

show abstract

“…Te frst one can be called CNN−LSTM [26,46], which indicates that the CNN is used to extract the local features of the signal frst and the LSTM is used to further process the extracted features. Te second one can be called LSTM−CNN [47,48], which means that the LSTM is frst used for the extraction of the overall features of the signal, and the CNN is subsequently used for further processing of the signal. Te third type can be called CNN+LSTM [49,50], in which both CNN and LSTM are used for feature extraction of the original input signal.…”

Section: Integration Of Cnn and Lstmmentioning

confidence: 99%

Classification of External Vibration Sources through Data-Driven Models Using Hybrid CNNs and LSTMs

Liang

Liu

Kaewunruen

et al. 2023

Structural Control and Health Monitoring

View full text Add to dashboard Cite

Excessive external vibrations could affect the normal functioning and integrity of sensitive buildings such as laboratories and heritage buildings. Usually, these buildings are exposed to multiple external vibration sources simultaneously, so the monitoring and respective evaluation of the vibration from various sources is necessary for the design of targeted vibration mitigation measures. To classify the sources of vibration accurately and efficiently, the advanced hybrid models of the convolutional neural network (CNN) and long short-term memory (LSTM) network were built in this study, and the models are driven by the extensive data of external vibration recorded in Beijing, and the parametric studies reveal that the proposed optimal model can achieve an accuracy of over 97% for the identification of external vibration sources. Finally, a real-world case study is presented, in which external vibration monitoring was carried out in a laboratory and the proposed CNN+LSTM model was used to identify the sources of vibration in the monitoring so that the impact of vibration from each source on the laboratory was analyzed statistically in detail. The results demonstrate the necessity of this study and its feasibility for engineering applications.

show abstract

“…The implementation of the DL methods in remote sensing for LSE estimations has been considered in recent years. As an example, Ye et al (2021) developed a novel deep neural network by combining the long short-term memory (LSTM) network and convolutional neural network (CNN) for estimating LSE, LST, atmospheric transmittance, upward radiance, and downward radiance (Ye et al 2022). Their model was applied to a hyperspectral image and validated by ground measurement data.…”

Section: Introductionmentioning

confidence: 99%

“…Their model was applied to a hyperspectral image and validated by ground measurement data. Nevertheless, due to the varying influence of the atmosphere and the variability of emissivity on different wavelengths, more analysis and validation should be done to confirm their result (Ye et al 2022). While former studies have showed the ability of DL to address temperature and emissivity separation, this machine learning algorithm can be employed to establish relationships between LSE and different spectral bands, fundamentally modelling these relationships and enabling LSE retrieval independent of temperature.…”

Section: Introductionmentioning

confidence: 99%

SWIR based estimation of TIR emissivity of bare soil surfaces using deep conditional generative adversarial network in Landsat data

Ataei,

Momeni,

Monadjemi

2024

Preprint

View full text Add to dashboard Cite

Aims Land surface emissivity (LSE) is an important variable in soil studies. Although there are various remote sensing methods to estimate LSE, accurately predicting LSE still remains a major challenge. Typically, the correlation between LSE and the Visible Near Infrared bands is employed for LSE estimation. However, some studies have raised some concerns about this correlation, especially in bare soil areas. Therefore, it is necessary to conduct further investigation to determine if there exists a nonlinear relationship between the LSE and other spectral bands, which was not detected by simple linear correlation/regression. Methods In this study, firstly, a deep Auto-encoder (DAE) network has been used to investigate the correlation between LSE and other spectral bands. Subsequently, we have applied a Conditional Generative Adversarial Network (CGAN) to estimate the LSE. The proposed CGAN was trained using the Landsat and ECOSTRESS satellite datasets. The performance of the developed network was then compared with NDVI-based method on satellite/simulated-based bare soil pixels. Results For satellite data, the RMSE (Root Mean Squared Error) and correlation coefficient (R) between the estimated LSE using proposed CGAN and ECOSTRESS LSE are 0.005 and 0.97, respectively. For the simulated data, the RMSE and R between the estimated LSE and the simulated one are 0.01 and 0.92, respectively. Conclusion The results of the DAE show considerable relationship between the LSE and Short-Wave Infrared bands which not be seen using simple linear correlation.

show abstract

Simultaneous Estimation of Land Surface and Atmospheric Parameters From Thermal Hyperspectral Data Using a LSTM–CNN Combined Deep Neural Network

Cited by 10 publications

References 17 publications

Land Surface Temperature Estimation from Landsat-9 Thermal Infrared Data Using Ensemble Learning Method Considering the Physical Radiance Transfer Process

Land Surface Temperature Estimation from Landsat-9 Thermal Infrared Data Using Ensemble Learning Method Considering the Physical Radiance Transfer Process

Classification of External Vibration Sources through Data-Driven Models Using Hybrid CNNs and LSTMs

SWIR based estimation of TIR emissivity of bare soil surfaces using deep conditional generative adversarial network in Landsat data

Contact Info

Product

Resources

About