Document image segmentation using fuzzy classifier and the dual-tree DWT

In this paper, we propose a new method for textual areas extraction of an image using a fuzzy classifier and dual-tree discrete wavelet transform. We have extended our text extraction scheme for classification of document images into text, background, and picture components. Three class fuzzy classifiers and a morphological post-processing operation is used for this purpose. The proposed method shows better results compared to the previous wavelet-based document image segmentation algorithms.