SavedAdd your journal

Journals / Afyon Kocatepe Üniversitesi Fen ve Mühendislik Bilimleri Dergisi / 2022 / Cilt: 22 - Sayı: 6

Comparison of Performances of CNN, SVM and RF Algorithms in Classification of Hyperspectral Images Obtained from Different Platforms

Farklı Platformlardan Elde Edilen Hiperspektral Görüntülerin Sınıflandırılmasında Evrişimli Sinir Ağları, Destek Vektör Makineleri ve Rastgele Orman Algoritmalarının Performanslarının Karşılaştırılması

Eren Can SEYREK
Murat UYSAL
Pages
1368–1379
Publication Date
2022-12-28
DOI
—
🏛️Bibliographic metadata and full text access for this article are provided via DergiPark.
DergiPark Record ↗
✨

AI Summary & Key Takeaways

AI Insight
⚡ TL;DR(Too Long; Didn't Read)

When the classification performances were examined, it was seen that the highest overall accuracy values were obtained by CNN as 87.78%, 99.82% and 96.89% for the data sets, respectively.

📌 3 Key Takeaways

🎯
Objective & Hypothesis:Hyperspectral Images (HSI) are employed in many fields, owing to the high spectral resolution that offer.
📊
Key Findings & Data:In this regard, the performance of various Machine Learning (ML) algorithms in the classification of HSGs have been investigated.
💡
Conclusion & Impact:When the classification performances were examined, it was seen that the highest overall accuracy values were obtained by CNN as 87.78%, 99.82% and 96.89% for the data sets, respectively.
🔬Design:Modeling & Simulation
📍Scope:Comprehensive Sample
⚙️Variables:Hiperspektral görüntü · Görüntü sınıflandırma · Evrişimli sinir ağları

Abstract

Hyperspectral Images (HSI) are employed in many fields, owing to the high spectral resolution that offer. Classification of HSIs is a challenging process due to the high spectral resolution of the images. In this regard, the performance of various Machine Learning (ML) algorithms in the classification of HSGs have been investigated. Especially, Convolutional Neural Networks (CNN) architectures, have been specially developed for the classification of HSIs. Due to the high cost of hyperspectral imaging instruments, obtaining HSI datasets is challenging. In recent years, the costs of new generation hyperspectral imaging systems developed for manned and unmanned aerial vehicles (UAV) have been decreasing, and it has become possible to obtain high spatial resolution and cost-effective HSIs. In this study, it is aimed to examine the performance of various ML algorithms in the classification of HSIs with different spectral and spatial resolutions obtained from various platforms. In this context, satellite-based HyRANK Loukia, manned aircraft-based Chikusei and UAV-based WHU-Hi HanChuan images were classified using Support Vector Machines, Random Forest and CNN algorithms. When the classification performances were examined, it was seen that the highest overall accuracy values were obtained by CNN as 87.78%, 99.82% and 96.89% for the data sets, respectively.

Özet

Hiperspektral Görüntüler (HSG), sağladığı yüksek spektral çözünürlük sayesinde birçok alanda kullanım alanına sahiptir. HSG’lerin sınıflandırılması, görüntülerin yüksek spektral çözünürlüğü sebebiyle zorlayıcı bir süreçtir. Bu bağlamda HSG’lerin sınıflandırılmasında birçok Makine Öğrenme (MÖ) algoritmasının performansı araştırılmıştır. Özellikle Derin Öğrenmenin alt dallarından biri olan Evrişimli Sinir Ağları (ESA) tabanlı birçok ağ mimarisi HSG’lerin sınıflandırılması için özel olarak geliştirilmiştir. Hiperspektral görüntüleme sistemlerinin (HGS) yüksek maliyetleri sebebiyle veri setlerinin elde edilmesi zordur. Son yıllarda insanlı ve insansız hava araçları (İHA) için geliştirilen yeni nesil hiperspektral görüntüleme sistemlerinin maliyetleri giderek düşmekte olup yüksek mekânsal çözünürlüklü ve uygun maliyetli HSG elde edilmesi mümkün hale gelmiştir. Bu çalışmada çeşitli platformlardan elde edilmiş farklı spektral ve uzamsal çözünürlükteki HSG’lerin sınıflandırılmasında çeşitli MÖ algoritmalarının performansının incelenmesi amaçlanmıştır. Bu kapsamda uydu tabanlı HyRANK Loukia, hava aracı tabanlı Chikusei İHA tabanlı WHU-Hi HanChuan isimli görüntüler Destek Vektör Makineleri, Rastgele Orman ve ESA algoritmaları kullanılarak sınıflandırılmıştır. Sınıflandırma performansları incelendiğinde en yüksek genel doğruluk değerleri veri setleri için sırasıyla %87,78, %99,82 ve %96,89 olarak ESA tarafından elde edildiği görülmüştür.

🔖

Cite This Article

SEYREK, E. C., UYSAL, M. (2022). Comparison of Performances of CNN, SVM and RF Algorithms in Classification of Hyperspectral Images Obtained from Different Platforms. Afyon Kocatepe Üniversitesi Fen ve Mühendislik Bilimleri Dergisi, Cilt: 22 - Sayı: 6, 1368–1379. https://ajindex.com/articles/874202

📄 Full Text PDF Viewer

📄

PDF görüntüleyici hazırlanıyor...

References (7)

  1. Adao T., Hruska J., Padua L., Bessa J., Peres E., Morais R. et al., 2017. Hyperspectral imaging: A review on UAV-based sensors, data processing and applications for agriculture and forestry. Remote Sensing, 9(11), 1110. Akar O. and Tunc Gormus E., 2021. Land use/land cover mapping from airborne hyperspectral images with machine learning algorithms and contextual information. Geocarto International, 37(14), 3963-3990.
  2. Ardouin J.P., Levesque J. and Rea T.A., 2007. A demonstration of hyperspectral image exploitation for military applications, 2007 10th International Conference on Information Fusion, 1-8.
  3. Audebert N., Le Saux B. and Lefevre S., 2019. Deep learning for classification of hyperspectral data: A comparative review. IEEE Geoscience and Remote Sensing Magazine, 7(2), 159-173. Bhosle K. and Musande V., 2020, Evaluation of CNN model by comparing with convolutional autoencoder and deep neural network for crop classification on hyperspectral imagery. Geocarto International, 1(15), 813-827.
  4. Boser B.E., Guyon I.M. and Vapnik V.N., 1992. A training algorithm for optimal margin classifiers. Proceedings of the 5th Annual ACM Workshop on Computational Learning Theory, 144-152. Breiman L., 2001. Random Forests. Machine learning, 45(1), 5-32. Chan J.C.W. and Paelinckx D., 2008. Evaluation of Random Forest and Adaboost Tree-based Ensemble Classification and Spectral Band Selection for Ecotope Mapping Using Airborne Hyperspectral Imagery. Remote Sensing of Environment, 112(6), 2999-3011.
  5. Chen S., Jin M. and Ding J., 2020. Hyperspectral remote sensing image classification based on dense residual three-dimensional convolutional neural network. Multimedia Tools and Applications, 80(2), 1859-1882.
  6. Christovam L.E., Pessoa G.G., Shimabukuro M.H. and Galo M.L.B.T., 2019. Land use and land cover classification using hyperspectral imagery: evaluating the performance of spectral angle mapper, support vector machine and random forest. International Archives of the Photogrammetry, Remote Sensing & Spatial Information Sciences, 10–14 June 2019,
  7. Enschede, The Netherlands, 1841-1847. Congalton R.G. and Green K., 2019. Assessing the accuracy of remotely sensed data: principles and practices. CRC Press, 328, Boca Raton, FL. Crucil G., Castaldi F., Aldana-Jague E., van Wesemael B., Macdonald A. and Van Oost K., 2019. Assessing the performance of UAS-compatible multispectral and hyperspectral sensors for soil organic carbon prediction. Sustainability, 11(7), 1889. Erturk A., Iordache