Band Selection for Hyperspectral Image Classification by a Sliding Window Model

摘要

We investigate how to better use mutual information (MI) to select bands for hyperspectral image classification with less human intervention. Mutual information effectively measures the statistical dependence between two random variables. By modeling ground truth (e.g., a reference map) as one of the two random variables, MI can be used to find the spectral bands that contribute most to image classification. Extending our earlier work, we propose a sliding window model and apply mutual information to construct the estimated reference map, which need less human intervention. Experiments on the AVIRIS 92AV3C data set show that the proposed approach outperformed the benchmark methods, removing up to 55% of bands without significant loss of classification accuracy, compared to the 40% from that using the reference map accompanied with the data set. Meanwhile, its performance is found to be much robust to accuracy degradation when bands are cut off beyond 60%, revealing a better agreement in the mutual information estimation.