Prior research in scene classification has shown that high-level information can be inferred from low-level image features. Classification rates of roughly 90% have been reported using low-level features to predict indoor scenes vs. outdoor scenes. However, the high classification rates are often achieved by using computationally expensive, high-dimensional feature sets, thus limiting the practical implementation of such systems. We show that a more computationally efficient approach to indoor/outdoor classification can yield classification rates comparable to the best methods reported in the literature. A low complexity, low-dimensional feature set is used in conjunction with a two-stage support vector machine classification scheme to achieve a classification rate of 90.2% on a large database of consumer photographs.
Department, Program, or Center
Chester F. Carlson Center for Imaging Science (COS)
Pattern Recognition 4 (2002) 146-149
RIT – Main Campus