Remote Sensing Image Classification Based on Fuzzy Associative Classification

The classification of remote sensing images is one of the most important issues in the remote sensing field. Due to inherent variability and uncertainty of the data, training data is hard to obtain in most real-world applications, which impact the classification accuracy of traditional classifiers greatly. In this paper, a novel fuzzy associative classifier based on fuzzy association rules, namely fuzzy associative remote sensing classification (FARSC), is developed for the classification of remote sensing images. The proposed algorithm employs fuzzy C-means to partition quantitative attributes according to their intrinsic characteristics, adopts new jointing and pruning techniques without generating useless candidate itemsets, and introduces a weighted parameter to score the fuzzy association rules, which fuses multiple rules to avoid the bias towards some classes. To evaluate the performance of the proposed algorithm, an experiment on the remote sensing image of Zhalong Nature Reserve is performed, compared with two other image classification algorithms: support vector machine and extreme leaning machine. The experimental results show that the proposed algorithm not only has higher classification accuracy, but also is insensitive to the variation of amount of the training data. Hence FARSC can effectively overcome the problem of the lack of training data set in the actual remote sensing classification and get a high classification accuracy.