Abstract
Although PCA has been widely used for feature extraction and data reduction, it suffers from three main drawbacks: high computational cost, large memory requirement and low efficacy in processing large datasets such as HSI. This column analysed two variations of PCA, namely SPCA and Seg-PCA. Seg-PCA can further improve classification accuracy whilst significantly reducing the computational cost and memory requirement, without requiring prior knowledge. There is potential to apply similar feature extraction and data reduction techniques in application areas beyond HSI when analysis of large dimensional datasets is required such as magnetic resonance imaging (MRI) and digital video processing.
Original language | English |
---|---|
Pages (from-to) | 149-154 |
Number of pages | 6 |
Journal | IEEE Signal Processing Magazine |
Volume | 31 |
Issue number | 4 |
DOIs | |
Publication status | Published - Jul 2014 |
Keywords
- hyperspectral Imaging
- feature extraction
- data reduction
- remote sensing
- covariance matrices
- hypercubes
- memory management
- principal component analysis
Fingerprint
Dive into the research topics of 'Effective feature extraction and data reduction with hyperspectral imaging in remote sensing'. Together they form a unique fingerprint.Prizes
-
IET V&I Best PhD Thesis Award (only one in UK per year)
Zabalza, J. (Recipient), Ren, J. (Recipient) & Marshall, S. (Recipient), Dec 2016
Prize: Prize (including medals and awards)
File