Support Vector Machine Based Classification for Hyperspectral Remote Sensing Images after Minimum Noise Fraction Rotation Transformation

2011 ◽  
Author(s):  
Zhang Denghui ◽  
Yu Le
Keyword(s):  
Remote Sensing ◽  
Support Vector Machine ◽  
Hyperspectral Remote Sensing ◽  
Support Vector ◽  
Remote Sensing Images ◽  
Minimum Noise Fraction ◽  
Rotation Transformation ◽  
Hyperspectral Remote Sensing Images ◽  
Noise Fraction ◽  
Minimum Noise

Mapping lithology in Canada's Arctic: application of hyperspectral data using the minimum noise fraction transformation and matched filtering

2005 ◽  
Vol 42 (12) ◽  
pp. 2173-2193 ◽  
Author(s):  
J R Harris ◽  
D Rogge ◽  
R Hitchcock ◽  
O Ijewliw ◽  
D Wright
Keyword(s):  
Remote Sensing ◽  
Hyperspectral Remote Sensing ◽  
Hyperspectral Data ◽  
Matched Filtering ◽  
Lithological Mapping ◽  
Minimum Noise Fraction ◽  
Logical Mapping ◽  
End Members ◽  
Noise Fraction ◽  
Minimum Noise

A test site in southern Baffin Island, Canada has been established to study the applications of hyperspectral data to lithological mapping. Good bedrock exposure and minimal vegetation cover provide an ideal environment for the evaluation of hyperspectral remote sensing. Airborne PROBE hyperspectral data were collected over the study site during the summer of 2000. Processing methods involved (1) applying a minimum noise fraction (MNF) transformation to the data and visual interpretation of a ternary colour MNF image to produce a lithological–compositional map, and (2) selection of end members from the MNF image followed by matched filtering based on the selected end members to produce a lithological–compositional map. Both methods have produced a lithological map that compares favourably with the existing geological map. Although lichen imparts a similarity to the spectra throughout the visible and near infrared and short-wave infrared ranges, this study has shown that enough variability in the spectra as a function of different mineralogy was present to successfully discriminate one major lithological group (metatonalites) and three compositional units (psammites, quartzites, and monzogranites). Vegetation could be clearly distinguished, which in this area only is a good proxy for mapping metagabbroic rocks. Furthermore, discrimination of slightly different compositional units within the psammites and the metatonalites was also possible. The results from this study indicate that hyperspectral remotely sensed imagery holds promise for lithological mapping in Canada's North, although further analysis is required in different geologic environments in Canada's North to validate hyperspectral remote sensing as a useful aid to litho logical mapping.

scholarly journals Optimized Lithological Mapping from Multispectral and Hyperspectral Remote Sensing Images Using Fused Multi-Classifiers

2020 ◽  
Vol 12 (1) ◽  
pp. 177 ◽  
Author(s):  
Mahendra Pal ◽  
Thorkild Rasmussen ◽  
Alok Porwal
Keyword(s):  
Remote Sensing ◽  
Kappa Coefficient ◽  
Hyperspectral Remote Sensing ◽  
Support Vector ◽  
Landsat 8 ◽  
Remote Sensing Images ◽  
Spectral Angle Mapper ◽  
Lithological Mapping ◽  
Accuracy Index ◽  
Hyperspectral Remote Sensing Images

Most available studies in lithological mapping using spaceborne multispectral and hyperspectral remote sensing images employ different classification and spectral matching algorithms for performing this task; however, our experiment reveals that no single algorithm renders satisfactory results. Therefore, a new approach based on an ensemble of classifiers is presented for lithological mapping using remote sensing images in this paper, which returns enhanced accuracy. The proposed method uses a weighted pooling approach for lithological mapping at each pixel level using the agreement of the class accuracy, overall accuracy and kappa coefficient from the multi-classifiers of an image. The technique is implemented in four steps; (1) classification images are generated using a variety of classifiers; (2) accuracy assessments are performed for each class, overall classification and estimation of kappa coefficient for every classifier; (3) an overall within-class accuracy index is estimated by weighting class accuracy, overall accuracy and kappa coefficient for each class and every classifier; (4) finally each pixel is assigned to a class for which it has the highest overall within-class accuracy index amongst all classes in all classifiers. To demonstrate the strength of the developed approach, four supervised classifiers (minimum distance (MD), spectral angle mapper (SAM), spectral information divergence (SID), support vector machine (SVM)) are used on one hyperspectral image (Hyperion) and two multispectral images (ASTER, Landsat 8-OLI) for mapping lithological units of the Udaipur area, Rajasthan, western India. The method is found significantly effective in increasing the accuracy in lithological mapping.

Sign in / Sign up

Export Citation Format

Share Document