Fractal-based Ensemble Classification System for Hyperspectral Images

• Published in: IEEE geoscience and remote sensing letters Vol. 20; p. 1
• Main Authors: Beirami, Behnam Asghari, Pirbasti, Mehran A, Akbari, Vahid
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 01.01.2023; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Accuracy; Classification; Cube texture; Discriminant analysis; Ensemble learning; Feature extraction; Fractal analysis; Fractal dimension; Fractals; Hyperspectral image; Hyperspectral imaging; Image classification; Image enhancement; Iron; Mathematical models; Principal component analysis; Principal components analysis; Support vector machines; Training; Voting-based fusion
• ISSN: 1545-598X; 1558-0571
• DOI: 10.1109/LGRS.2023.3330608

According to the literature, the utilization of spatial features can significantly enhance the accuracy of hyperspectral image (HSI) classification. Fractal dimension (FD) features are powerful measures of texture, representing the local complexity of an image. In HSI classification, textural features are typically extracted from dimensionally reduced datacubes, such as principal component analysis (PCA). However, the effectiveness of textures obtained from alternative feature extraction methods in improving classification accuracy has not been extensively investigated. This study introduces a new ensemble support vector machine classification system that combines spectral features derived from PCA, minimum noise fraction, linear discriminant analysis, and FD features derived from these feature extraction methods. The final results on two HSI datasets, namely Indian Pines and Pavia University, demonstrate that the proposed classification method achieves approximately 95.75% and 99.36% accuracies, outperforming several other spatial-spectral HSI classification methods.