A comparison of individual and combined classification methods for processing of ASTER data

A comparison of individual and combined classification methods for processing of ASTER data

Mineral mapping using remote sensing techniques has been developed for application in the field of deposit exploration. However, the classification methodology in the processing of satellite images is one of the most important steps that affects the accuracy of resultant maps. An integration use of several classifiers is a helpful strategy to achieve more accurate results. The study aims to classify an ASTER scene with the use of Hyperion unmixing results of Lahrud, Iran. The detected minerals by Mixture Tuned Matched Filtering (MTMF) method on Hyperion image were used as training classes. The separability score was computed between classes imported from Hyperion data analysis. In order to improve the accuracy of upcoming processes, classes with high similarity (with low separability) were combined. Therefore, 6 classes with adequate separability score were determined as final classes for classification. The classification of ASTER scene was then performed with the use of four individual and four combined classifiers. An accuracy analysis was performed to compare the functionality of each classifier and the Max rule method demonstrated the best performance among all classifiers tested in this study regarding to its highest overall accuracy.

Mineral mapping using remote sensing techniques has been developed for application in the field of deposit exploration. However, the classification methodology in the processing of satellite images is one of the most important steps that affects the accuracy of resultant maps. An integration use of several classifiers is a helpful strategy to achieve more accurate results. The study aims to classify an ASTER scene with the use of Hyperion unmixing results of Lahrud, Iran. The detected minerals by Mixture Tuned Matched Filtering (MTMF) method on Hyperion image were used as training classes. The separability score was computed between classes imported from Hyperion data analysis. In order to improve the accuracy of upcoming processes, classes with high similarity (with low separability) were combined. Therefore, 6 classes with adequate separability score were determined as final classes for classification. The classification of ASTER scene was then performed with the use of four individual and four combined classifiers. An accuracy analysis was performed to compare the functionality of each classifier and the Max rule method demonstrated the best performance among all classifiers tested in this study regarding to its highest overall accuracy.