Satellite image classification using a classifier integration model

• Published in: 2011 9th IEEE/ACS International Conference on Computer Systems and Applications (AICCSA) pp. 90 - 94
• Main Authors: Dong-Chul Park, Taekyung Jeong, Yunsik Lee, Soo-Young Min
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.12.2011
• Subjects: Accuracy; classification; classifier fusion; Computer integrated manufacturing; Discrete cosine transforms; Feature extraction; image data; local classifier; Satellites; Training data; Vectors
• ISBN: 9781457704758; 1457704757
• ISSN: 2161-5322
• DOI: 10.1109/AICCSA.2011.6126608

A new satellite image classification method using a classifier integration model(CIM)is proposed in this paper. CIM does not use the entire feature vectors extracted from the original data in a concatenated form to classify each datum, but rather uses groups of features related to each feature vector separately. In the training stage, a confusion table calculated from each local classifier that uses a specific feature vector group is drawn throughout the accuracy of each local classifier and then, in the testing stage, the final classification result is obtained by applying weights corresponding to the confidence level of each local classifier. The CIM is applied to the problem of satellite image classification on a set of image data. The results demonstrate that the CIM scheme can enhance the classification accuracy of individual classifiers that use specific feature vector group.