IEEE GEOSCIENCE AND REMOTE SENSING LETTERS 1 Multi-Source Composite Kernels for Urban Image Classification

Abstract — This work presents advanced classification methods for very high resolution images. Efficient multisource informa-tion, both spectral and spatial, is exploited through the use of composite kernels in support vector machine (SVM). Weighted summations of kernels accounting for separate sources of spectral and spatial information are analyzed and compared to classi-cal approaches such as pure spectral classification or stacked approaches using all the features in a single regressor. Model selection problems are addressed, as well as the importance of the different kernels in the weighted summation. Index Terms — Urban monitoring, very high resolution image, Support vector machines (SVM), multiple kernel learning. I