The week's pick
Reseach Article
Detection and Classification of Mahogany Tree Species via Satellite Imagery, Google Earth Engine, and Deep Learning
| International Journal of Computer Applications |
| Foundation of Computer Science (FCS), NY, USA |
| Volume 186 - Number 31 |
| Year of Publication: 2024 |
| Authors: Joy Roy, Muhammad Anwarul Azim, Abu Nowshed Chy, Mohammad Khairul Islam |
10.5120/ijca2024923822
|
Joy Roy, Muhammad Anwarul Azim, Abu Nowshed Chy, Mohammad Khairul Islam . Detection and Classification of Mahogany Tree Species via Satellite Imagery, Google Earth Engine, and Deep Learning. International Journal of Computer Applications. 186, 31 ( Jul 2024), 18-26. DOI=10.5120/ijca2024923822
Abstract
The mahogany tree is widely used in various industries worldwide for its valuable timber. In response to the increasing scarcity of mahogany wood, particularly in Bangladesh, this study presents an automated process for detecting and classifying mahogany tree species. Focusing on the University of Chittagong, Bangladesh region where multi-polygons delineating mahogany trees and other land cover types were generated using Google Earth Engine. Sentinel-2 satellite imagery from 2019 and 2020 provided spectral band wavelength data, enabling the creation of datasets with two classes: mahogany tree and non-mahogany tree cover types. To enhance model performance and accuracy, the four satellite datasets removed irrelevant columns and eliminated 7,661 duplicate entries, resulting in 2,339 unique entries. The refined dataset comprises 12 spectral band features and one class attribute, which significantly improves the classification accuracy. Supervised classification employed a multilayer perceptron deep neural network with three hidden layers. The model achieved a training accuracy of 99.76% and testing accuracy of 96.58%, with a precision of 98.36%, recall of 99.05%, and F1-Score of 97.6%. The optimal performance was observed with the Spectral band wavelength dataset of Satellite Surface Reflectance 2019 images using the Adam optimizer. This research contributes to advancing automated methods for mahogany tree detection, facilitating conservation efforts for tree species, and informing resource management practices in Bangladesh and beyond.
References
- J. F. Mongillo and L. Zierdt-Warshaw, “Encyclopedia of environmental science,” 2000.
- S. Bridgewater, “A natural history of Belize: Inside the Maya Forest,” 2012.
- S. H. Brown, B. Mason, and M. Gardener, “Mahogany: Swietenia mahagoni Family: Meliaceae,” gardeningsolutions.ifas.ufl.edu
- D. A. Friess, E. P. Kudavidanage, and E. L. Webb, “The digital globe is our oyster,” Front. Ecol. Environ., vol. 9, no. 10, pp. 542–542, Dec. 2011, doi: 10.1890/11.wb.029.
- D. Butler, “Virtual globes: the web-wide world,” Nature, vol. 439, no. 7078, pp. 776–778, Feb. 2006, doi: 10.1038/439776a.
- M. C. Hansen et al., “High-resolution global maps of 21st-century forest cover change,” Science, vol. 342, no. 6160, pp. 850–853, Nov. 2013, doi: 10.1126/science.1244693.
- “Blog.” https://ai.googleblog.com/2013/11/the-first-detailed-maps-of-global.html (accessed Jan. 22, 2021).
- D. Lu and Q. Weng, “A survey of image classification methods and techniques for improving classification performance,” Int. J. Remote Sens., vol. 28, no. 5, pp. 823–870, Mar. 2007, doi: 10.1080/01431160600746456.
- S. R. Phinn, “A framework for selecting appropriate remotely sensed data dimensions for environmental monitoring and management,” Int. J. Remote Sens., vol. 19, no. 17, pp. 3457–3463, Nov. 1998, doi: 10.1080/014311698214136.
- J. R. Jensen and D. C. Cowen, “Remote sensing of urban/suburban infrastructure and socio-economic attributes,” Photogramm. Eng. Remote Sens., vol. 65, pp. 611–622, 1999.
- D. Chen and D. Stow, “The effect of training strategies on supervised classification at different spatial resolutions,” Photogramm. Eng. Remote Sens., vol. 68, no. 11, pp. 1155–1162, 2002.
- Y. Sun, Y. Liu, G. Wang, and H. Zhang, “Deep Learning for Plant Identification in Natural Environment,” Comput. Intell. Neurosci., vol. 2017, p. 7361042, May 2017, doi: 10.1155/2017/7361042.
- T. Mizoguchi, A. Ishii, H. Nakamura, T. Inoue, and H. Takamatsu, “Lidar-based individual tree species classification using convolutional neural network,” Videometrics, Range Imaging, and Applications XIV. 2017, doi: 10.1117/12.2270123.
- K. Zhang and B. Hu, “Individual Urban Tree Species Classification Using Very High Spatial Resolution Airborne Multi-Spectral Imagery Using Longitudinal Profiles,” Remote Sensing, vol. 4, no. 6, pp. 1741–1757, Jun. 2012, doi: 10.3390/rs4061741.
- J. S. H. Lee, S. Wich, A. Widayati, and L. P. Koh, “Detecting industrial oil palm plantations on Landsat images with Google Earth Engine,” Remote Sensing Applications: Society and Environment, vol. 4. pp. 219–224, 2016, doi: 10.1016/j.rsase.2016.11.003.
- I. Goodfellow, Y. Bengio, A. Courville, and Y. Bengio, Deep learning, vol. 1. MIT Press Cambridge, 2016.
- Zhang, H. K., Roy, D. P., Yan, L., Li, Z., Huang, H., Vermote, E., Skakun, S., & Roger, J. C. (2018, September 1). Characterization of Sentinel-2A and Landsat-8 top of atmosphere, surface, and nadir BRDF adjusted reflectance and NDVI differences. Remote Sensing of Environment. https://doi.org/10.1016/j.rse.2018.04.031
Index Terms
Keywords