Wavelet based semisupervised learning for satellite image classification using ensemble projection

Image classification techniques play a significant role in the remote sensing imagery. Many of the researchers found some difficulties while doing the analysis of satellite images. During the classification task, many questions have arisen in the minds of the experts and they might face many challenging issues. SSEP (Semi Supervised Ensemble Projection) is a newly adopted method that yields better accuracy even when the satellite image dataset comprises of limited labeled data and great quantity of unlabeled data. Initially, it is common to extract the preliminary features like color, structure, textures for the given image. In this article, we have not only proposed a new Gaussian normal affinity to describe the nearest neighbor by ensemble process in an accurate way which ensures the reliability and diversity, but also we have applied wavelet transformation to texture feature for enhancing the accuracy of classification. Sparse coding technique was mainly meant to overcome the redundancy. The effectiveness of the proposed method is successfully illustrated by using high resolution satellite dataset.