Publication Type : Conference Paper
Publisher : 2017 International Conference on Communication and Signal Processing (ICCSP)
Source : 2017 International Conference on Communication and Signal Processing (ICCSP), IEEE, Chennai, India, p.1313 – 1317 (2017)
Url : https://ieeexplore.ieee.org/document/8286595
Keywords : ANN, ANN model, artificial materials, Artificial Neural Network, Artificial neural networks, Clouds, data analysis, ecology, GIS Processing, India, land cover, Land surface, Land surface temperature, LST, moisture, neural nets, Predictive models, Regression analysis, regression method, Remote sensing, Remotely sensed data, RMSE method, rural land cover area types, Salem district, Satellite data, satellite data analysis, Soil, Soil moisture analysis, spectral radiation, Tamilnadu, Temperature prediction, temperature sensors, urban land cover areas, vegetation, vegetation change parameters, Vegetation index, Vegetation mapping, Yercaud
Campus : Coimbatore
School : School of Engineering
Center : Computational Engineering and Networking
Department : Center for Computational Engineering and Networking (CEN)
Verified : Yes
Year : 2017
Abstract : Unexpected and unrestrained extension of urban land cover areas, rural land cover area types have been replaced with artificial materials. Such impact needs to consider the factor of increasing Land surface temperature. It is a necessary aspect to be considered in monitoring various global changes. The most efficient method to find LST is by means of satellite data analysis. In this paper, prediction of LST uses vegetation indices for analyzing vegetation change parameters and soil moisture analysis then the statistical outcome of the study area as the processing inputs exploiting an artificial neural network (ANN). Spectral radiation having the ability to penetrate through clouds, hence the vegetation indices and LST can be evaluated even under critical, cloudy conditions. The ANN model were trained and then tested for respective cases where LST was attainable. This model takes a series of earlier period LST values, identify with the outline of transform within the data and promote prediction or future time values. The performance analyses of the framework were estimated based on typical evaluation methods of RMSE and Regression. The best resultant model was utilized to predict LST where the observing circumstances were available. Conclusions affirm that ANN model serves good in prediction of LST with the real world observations. The proposed study using ANN design is useful in predicting LST in Yercaud, a hilly region in Salem district, of Tamilnadu using the remotely sensed data even under various circumstances of the ecology.
Cite this Research Publication : E. V. Shanmugapriya and Dr. Geetha Srikanth, “A framework for the prediction of land surface temperature using artificial neural network and vegetation index”, in 2017 International Conference on Communication and Signal Processing (ICCSP), Chennai, India, 2017, pp. 1313 – 1317.