Pattern Recognition of Satellite Imageries of Somwarpet Taluk of Kodagu District: Land Use Patterns Classification

Author: Vaibhav Chittora*, S.S. Patil, Hemamalini H.C., Pramit Pandit and Prabhat Kumar

Download PDF

Abstract

The land use, land cover classification of satellite imagery is compelling to analyses derives the knowledge. The current scenario playing vital role in pattern classification to recognize pattern behaviour through the machine learning algorithms are utilized for pattern recognition and their performance. The imagery obtained by Sentinel-2 Satellite on February 2018 for Somwarpet Taluk, Kodagu District (Karnataka) using ERDAS IMAGINE image processing. In order to classify land cover types, training is needed to create a set of statistics that describes the spectral response patterns of each type of land cover draw the features, the quality of training feature set ensures the success of classification in accuracy nearer to ground truth. Maximum Likelihood Classification, Minimum Distance to Mean Classification, Mahalanobi’s Distance Classification and Spectral Correlation Mapper Classification were outperformed algorithms. Accuracy of the classification of data set and classifier were

Keywords

Accuracy assessment, Classification algorithm, Kappa statistic, ERDAS Imagine.

Conclusion

This study of the Somwarpet Taluk of Kodagu District Karnataka, India demonstrates that the utilization of spatial multi-transient satellite imagery with the guide of GIS and RS innovation can assume a fundamental part in computing spatial and temporal phenomena, previously it was not possible through traditional digital planning. In this study seven LULC classes were classified in the study area namely Agricultural Crops, Plantation Crops, Built up, Forest, Barren land, Scrub land and Water bodies. In the classification phase four supervised classification algorithms were deployed to classify the image. The four algorithms are maximum Likelihood classification algorithm, Minimum Distance, Mahalanobis Distance and Spectral Correlation Mapper were performed to the image. Several measures of classification accuracy were evaluated in this study, namely overall accuracy, kappa coefficient and f measures. Many measures of classification accuracy may be derived from a confusion m

References

-

How to cite this article

Vaibhav Chittora, S.S. Patil, Hemamalini H.C., Pramit Pandit and Prabhat Kumar (2022). Pattern Recognition of Satellite Imageries of Somwarpet Taluk of Kodagu District: Land use Patterns Classification. Biological Forum – An International Journal, 14(