Wavelet based co-occurrence histogram features for texture classification with an application to script identification in a document image

In this paper, we propose a novel texture feature extraction method based on the co-occurrence histograms of wavelet decomposed images, which capture the information about relationships between each high frequency subband and that in low frequency subband of the transformed image at the corresponding level. The correlation between the subbands at the same resolution exhibits a strong relationship, indicating that this information is significant for characterizing a texture. The classification performance is tested on a set of 32 Brodatz textures using the different wavelet filter banks for the proposed feature set. The results are compared with those obtained by using the Gabor filters and the method proposed by Montiel et al. [Montiel, E., Aguado, A.S., Nixon, M.S., 2005. Texture classification via conditional histograms. Pattern Recognition Lett. 26, 1740–1751]. The proposed and the Gabor features are then used in the identification of the script of a machine printed document. The scheme has been tested on eight Indian language scripts including English. It is found to be robust to the skew generated in the process of scanning a document. The experiments are also performed on the images with orientations of different angles and with varying coverage of text. The classification performance is analyzed using the k-NN classifier. The experimental results demonstrate the effectiveness of the proposed texture features in achieving the improved classification performance.