Full Text

Abstract

This paper aims to perform satellite image processing using Machine Learning models and evaluate their prediction scores. This research will classify satellite images into four distinct categories, namely \"green area,\" \"desert,\" \"water,\" and \"cloudy, and will train and evaluateseveral criteria to highlight the model\"s exceptional performance, including precision, recall, F1-scores, and total accuracy. The model exhibits exceptional accuracy in correctly predicting and identifying positive instances, as seen by the near-perfect scores achieved for precision and recall across most classes. The F1 scores demonstrate a cohesive equilibrium across various measures, indicating the approach\"s efficacy. Significantly, the model has a remarkable overall accuracy rate of 99%, emphasizing its proficiency in precise picture categorization. The use of macro and weighted averages highlights the resilience and uniformity of its performance, irrespective of variations in class distribution. The findings presented in this study provide evidence supporting the appropriateness of the model for a range of applications, with a particular emphasis on computer vision and machine learning. Evaluation measures like accuracy, recall, and F1-score provide a detailed analysis of the model\"s capabilities, rendering them essential for assessing classification models.

References

1. Esmaeilzehi, A., Ahmad, M. O., & Swamy, M. (2021, November). SRNHARB: A deep lightweight image super-resolution network using hybrid activation residual blocks. Signal Processing: Image Communication, 99, 116509. https://doi.org/10.1016/j.image.2021.116509

2. Lu, T., Wang, J., Zhang, Y., Wang, Z., & Jiang, J. (2019, July 4). Satellite Image Super-Resolution via Multi-Scale Residual Deep Neural Network. Remote Sensing, 11(13), 1588. https://doi.org/10.3390/rs11131588

3. Wang, L., & Sun, Y. (2022, March 10). Image classification using a convolutional neural network with wavelet domain inputs. IET Image Processing, 16(8), 2037–2048. https://doi.org/10.1049/ipr2.12466

4. Choi, S. P. (2017, March 15). Extraction of Protein-Protein Interactions based on Convolutional Neural Network (CNN). KIISE Transactions on Computing Practices, 23(3), 194–198. https://doi.org/10.5626/ktcp.2017.23.3.194

5. Nandini, B. (2021, July 10). Detection of Skin Cancer using Inception V3 And Inception V4 Convolutional Neural Network (CNN) For Accuracy Improvement. RevistaGestãoInovação E Tecnologias, 11(4), 1138–1148. https://doi.org/10.47059/revistageintec.v11i4.2174

6. Park, J., Kim, H., & Paik, J. (2021, April 20). CF-CNN: Coarse-to-Fine Convolutional Neural Network. Applied Sciences, 11(8), 3722. https://doi.org/10.3390/app11083722

7. CHANDRASEKARAN, A. and KALLA, D. (2023) Heart disease prediction using chi-square test and linear regression.? Computer Science & Information Technology, 13, pp. 135-146.

8. Kalla, D., Samaah, F., Kuraku, S. & Smith, N. Phishing Detection Implementation Using Databricks and Artificial Intelligence. SSRN Electronic Journal 185, doi:10.2139/ssrn.4452780 (2023).

9. Wang, W., Zhu, M., Chi, X., & Xu, H. (2021, June). S-CNN-ESystem: An end-to-end embedded CNN inference system with low hardware cost and hardware-software time-balancing. Journal of Systems Architecture, p. 116, 102122. https://doi.org/10.1016/j.sysarc.2021.102122

10. Masoudi, B., &Danishvar, S. (2022). Deep multi-modal schizophrenia disorder diagnosis via a GRU-CNN architecture. Neural Network World, 32(3), 147–161. https://doi.org/10.14311/nnw.2022.32.009

11. Fabricius, D. (2016, November 16). Architecture before architecture: Frei Otto’s ‘Deep History.’ The Journal of Architecture, 21(8), 1253–1273. https://doi.org/10.1080/13602365.2016.1254667

12. Kim, J. (2022, October 17). Making Architecture Relevant to Underserved Communities: Mapping Reconsidered. Architecture, 2(4), 637–659. https://doi.org/10.3390/architecture2040034

13. Gallaher, J. (2020). Division (Architecture 5), and: \"Adopt\" as in \"a Program\" (Architecture 11), The Second Idea of Absence (Architecture 12), Nest (Architecture 17), and Nest III (Architecture 102). The Missouri Review, 43(4), 41–47. https://doi.org/10.1353/mis.2020.0048

14. Rani, P. A. S., & Singh, N. (2022, August 8). Paddy Leaf Symptom-based Disease Classification Using Deep CNN with ResNet-50. International Journal of Advanced Science Computing and Engineering, 4(2), 88–94. https://doi.org/10.30630/ijasce.4.2.83

15. Loey, M., Manogaran, G., Taha, M. H. N., & Khalifa, N. E. M. (2021, February). Fighting against COVID-19: A novel deep learning model based on YOLO-v2 with ResNet-50 for medical face mask detection. Sustainable Cities and Society, p. 65, 102600. https://doi.org/10.1016/j.scs.2020.102600

16. Kalla, D., & Chandrasekaran, A. (2023). HEART DISEASE PREDICTION USING MACHINE LEARNING AND DEEP LEARNING. International Journal of Data Mining & Knowledge Management Process (IJDKP), 13(3). https://doi.org/10.5121/ijdkp.2023.13301.

17. Tchórzewski, J., &Wielgo, A. (2021, December 23). A neural model of human gait and its implementation in MATLAB and Simulink Environment using Deep Learning Toolbox. Studia Informatica, 25. https://doi.org/10.34739/si.2021.25.03

18. McKenzie, A. L., Muñoz, C. X., & Armstrong, L. E. (2015, December 1). Accuracy of Urine Color to Detect Equal to or Greater Than 2% Body Mass Loss in Men. Journal of Athletic Training, 50(12), 1306–1309. https://doi.org/10.4085/1062-6050-51.1.03

19. Stamoulis, D. T., & Hauenstein, N. M. (1993, December). Rater training and rating accuracy: Training for dimensional accuracy versus training for rate differentiation. Journal of Applied Psychology, 78(6), 994–1003. https://doi.org/10.1037/0021-9010.78.6.994

Keywords

Artificial Intelligence, Azure Databricks, Machine Learning, Deep Learning, Convolution Neural Network, Residual Network, Artificial Neural Network, ResNet 50, Image Processing, Vikram lander, Pragyan Lunar Rover, Perseverance Mars Rover, LRV, NASA, Remote Sensing and Satellite Image Processing.