Abstract

This study uses the UCI urban land cover dataset to propose a unique integrated discriminant model. Computers are used in remote sensing image classification to process features and classify models of images captured by satellite remote sensing. Current research primarily focuses on feature extraction from images and model training to accomplish feature-based classification for accurate land type classification. This study suggests a novel approach for the latter by training each set of data taken from the many coarser scales of the original image using the supervised machine learning algorithms Random Forest, SVM, Naïve Bayes, and KNN. The prediction accuracy of the algorithms is then compared. The model at that scale is the algorithm that works best with the matching scale data. Lastly, the ultimate result is obtained by weighing the predictions made by each model. Tests have demonstrated that the model outperforms current techniques on every scale.

References

1. Johnson, B., Xie, Z., 2013. Classifying a high-resolution image of an urban area using super-object information. ISPRS Journal of Photogrammetry and Remote Sensing, 83, 40-49.
2. Johnson, B., 2013. High resolution urban land cover classification using a competitive multi-scale object-based approach. Remote Sensing Letters, 4 (2), 131-140.
3. Hegde, Gaurav et al. 'Urban Land Cover Classification Using Hyperspectral Data.' ISPRS Technical Commission VIII Symposium, translated by Isprs Tech Commission et al., vol. 40-8, 2014
4. 'Urban Land Cover Classification Using Hyperspectral Data.' ISPRS Technical Commission VIII Symposium, translated by Isprs Tech Commission et al., vol. 40-8, 2014
5. Tong, Xiaohua et al. 'Urban Land Cover Classification with Airborne Hyperspectral Data: What Features to Use?' Ieee Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol. 7, no. 10, 2014, pp. 3998-4009, doi:10.1109/jstars.2013.2272212.
6. Li, Su et al. 'A Novel Approach Based on the Combination Image of Fraction Image and Normalized Mnf Image to Urban Land Use/Cover Mapping.' IEEE International Geoscience and Remote Sensing Symposium (IGARSS), translated by Ieee, 2007
7. Li, Xinwu et al. 'Urban Land Cover Classification with High-Resolution Polarimetric Sar nterferometric Data.' Canadian Journal of Remote Sensing, vol. 36, no. 3, 2010, pp. 236-47, doi:10.5589/m10-046.
8. Lu, Dengsheng et al. 'Land Cover Classification in a Complex Urban-Rural Landscape with Quickbird Imagery.' Photogrammetric Engineering and Remote Sensing, vol. 76, no. 10, 2010, pp. 1159-68, doi:10.14358/pers.76.10.1159.
9. Novack, T. and H. J. H. Kux. 'Urban Land Cover and Land Use Classification of an Informal settlement Area Using the Open-Source Knowledge-Based System Interimage.' Journal of Spatial Science, vol. 55, no. 1, 2010, pp. 23-41, doi:10.1080/14498596.2010.487640.
10. Qiu, Xiaomin et al. 'Incorporating Road and Parcel Data for Object-Based Classification of Detailed Urban Land Covers from Naip Images.' Giscience & Remote Sensing, vol. 51, no. 5, 2014, pp. 498-520, doi:10.1080/15481603.2014.963982.
11. Zheng, Gan et al. A Random Forest Based Method for Urban Object Classification Using Lidar Data and Aerial Imagery. 2015. 2015 23rd International Conference on Geoinformatics. Proceedings.
12. Wentz, Elizabeth A. et al. 'Land Use and Land Cover Mapping from Diverse Data Sources for an and Urban Environments.' Computers Environment and Urban Systems, vol. 30, no. 3, 2006, pp. 320-46, doi:10.1016/j.compenvurbsys.2004.07.002.
13. https://scholar.google.com/citations?user=xh9HeqgAAAAJ&hl=en
14. https://scholar.google.com/citations?user=n8yVDWQAAAAJ&hl=en
15. https://scholar.google.com/citations?user=MOfCYLwAAAAJ&hl=en

Download PDF
Back