Comparison of Pca (Principal Component Analysis) and Band Ratio Methods Using Landsat 9 Images for Land Cover Visualization Mapping
Case Study: Lubuk Begalung District, Padang City
Abstract
Identification of land cover using satellite imagery through supervised and unsupervised classifications has been successfully carried out in various locations. This study aims to assess the accuracy of Landsat-9 OLI imagery by identifying land cover by observant the spectral responses of settlements, industrial buildings, rice fields, forests, wetlands, and open land using Principle Component Analysis (PCA) and Band Ratio methods. The success of land cover mapping was evaluated statistically using a confusion matrix. The acceptable level of accuracy is 85% with a Cohen’s kappa coefficient greater than 0.8. The results showed that the PCA and Band Ratio methods were not successful in mapping land cover at the study site because the accuracy level of both was at the specified moderate level. PCA has 80.5% accuracy with 0.75 kappa better than Band Ratio (80.5% accuracy, 0.76 kappa). The level of spectral sensitivity in industrial buildings and open land on PCA is much better than Band Ratio. However, in the residential and rice fields the Band Ratio is much better at detecting than PCA.
Downloads
References
[2] A. Singh, “Review Article Digital change detection techniques using remotely-sensed data,” vol. 1161, 2010, doi: 10.1080/01431168908903939.
[3] L. M. Prayogo and A. Basith, “Comparison of Robert's Filter Methods, Segmentation and Band Ratio on Landsat 8 Imagery for Coastline Analysis," no. December, 2021, doi: 10.21107/rekayasa.v14i3.10300.
[4] R. M. McCoy, Field Methods in Remote Sensing. 72 Spring Street, New York, NY 10012: The Guilford Press, 2005.
[5] J. G. Masek et al., “Landsat 9: Empowering open science and applications through continuity,” Remote Sens. Environ., vol. 248, no. May, 2020, doi: 10.1016/j.rse.2020.111968.
[6] N. Grumman, “Mission Management and Operations: Spacecraft Provider and Observatory Integration,” 2019.
[7] J. A. Richards and X. Jia, “Remote sensing digital image analysis: An introduction,” Remote Sens. Digit. Image Anal. An Introd., no. January 2006, pp. 1–439, 2006, doi: 10.1007/3-540-29711-1.
[8] B. Raharja, A. Setianto, A. D. Titisari, and M. Resources, “APPLICATION OF PRINCIPLE COMPONENT ANALYSIS IN THE MAPPING OF HYDROTHERMAL ALTERATION USING LANDSAT 8 IMAGE IN KOKAP , KULON PROGO,” no. September, pp. 873–893, 2019.
[9] Y. L. Æ. L. Wang, Æ. W. U. C. Æ. S. Peng, and J. T. T. Æ. Y. H. Æ. H. H. Æ, “Ratiometric imaging of gastrodermal lipid bodies in coral – dinoflagellate endosymbiosis,” 2005.
[10] P. Wicaksono and P. Danoedoro, “Multitemporal Vegetation Cover Analysis Vegetation Cover Mapping Using Alos Avnir-2: the Importance of Atmospheric Effect Normalization on Multitemporal Analysis,” no. February, pp. 104–107, 2012, doi: 10.13140/2.1.5030.6887.
[11] S. Zuhannisa, B. E. Cahyono, and N. Priyantari, “Utilization of Landsat 8 Imagery for Mapping the Potential of Gold Mineralization in South Tapanuli Regency, North Sumatra,” no. July 2019, 2021, doi: 10.13057/ijap.v9i01.25478.
[12] H. Wulansari, “Land Use Classification Accuracy Test Using Maximum Likelihood Defuzzification Method Based on Alos Avnir-2 Citra Image,” BHUMI J. Agrar. dan Pertanah., vol. 3, no. 1, p. 98, 2017, doi: 10.31292/jb.v3i1.96.
