Image classification by exploiting the spatial context information

Finding the effective image representation is an important problem for classification. Previous approaches have demonstrated the utility of the bag-of-feature (BoF) models. These methods are interesting due to the computational efficiency and conceptual simplicity. However, it is achieved by discarding the spatial context information. Furthermore, it may suffer from the quantization error introduced by the hard quantization of local features. To address these issues, we proposed an effective image representation that exploits the spatial context information. Specifically, the visual codebook is constructed on the pair-wise descriptors lied in spatial neighborhoods which can capture the near-context information, and the spatial pyramid structure is further combined to capture the far-context information. Then for image classification, an effective soft quantization method is proposed, which can accurately represent the original features by the regression of neighboring visual words. To evaluate the effectiveness of the proposed method, we compared it with existing BoF representations on benchmark datasets including Scenes-15 and Caltech 101 in image classification. The experimental results demonstrate the superiority of the proposed method compared with state-of-the-art methods.