Text Extraction and Document Image Segmentation Using Matched Wavelets and MRF Model

Abstract

In this paper, we have proposed a novel scheme for the extraction of textual areas of an image using globally matched wavelet filters. A clustering-based technique has been devised for estim ating globally matched wavelet filters using a collection of groundtruth images. We have extended our text extraction scheme for the segmentation of document images into text, background, and picture components (which include graphics and continuous tone images). Multiple, two-class Fisher classifiers have been used for this purpose. We also exploit contextual information by using a Markov random field formulation-based pixel labeling scheme for refinement of the segmentation results. Experimental results have established effectiveness of our approach.

Structure detection and segmentation of documents using 2D stochastic context-free grammars

“…Nevertheless, there are many methods in the literature that provide approximate solutions to the problem. One example is the Graph Cut algorithm [15] used in many segmentation tasks [29]. The method relies on the fact that many computer vision problems can be formulated in terms of an energy minimization function, and provides a local minimum of the minimization function based on the most likely cut in the graph.…”

An Improved Gray Multi-focus Image Fusion Algorithm Based on Dual-Tree Complex Wavelet Transform

“…"If-then" is a conditional statement that contains fuzzy logic, In its simplest form "if-then" strategy is: If x is A then y is B, it means that if x is A, then y is B, A and B are fuzzy linguistic variables, the relationship between A and B is the fuzzy relationship, denoted as A → B, "X is A" is a prerequisite, "y is B" is the conclusion [7][8] . Process applications "if-then" reasoning strategies can be divided into the following three steps: Discuss all possible preconditions.…”

Document Content Extraction Using Automatically Discovered Features

“…In [13], Nicolas and Dardenne et al adapt and apply conditional random fields (CRF) to document image segmentation; they classifies 3x3 regions. Kumar and Gupta et al in [11] use globally matched wavelet filters to discriminate text from nontext within color document images; they classify individual pixels.…”