Complex documents images segmentation based on steerable pyramid features

Multiscale Region Projection Method to Discriminate Between Printed and Handwritten Text on Registration Forms

Techniques to identify printed and handwritten text in scanned documents differ significantly. In this paper, we address the question of how to discriminate between each type of writing on registration forms. Registration-form documents consist of various type zones, such as printed text, handwriting, table, image, noise, etc., so segmenting the various zones is a challenge. We adopt herein an approach called “multiscale-region projection” to identify printed text and handwriting. An important aspect of our approach is the use of multiscale techniques to segment document images. A new set of projection features extracted from each zone is also proposed. The classification rules are mining and are used to discern printed text and table lines from handwritten text. The proposed system was tested on 11[Formula: see text]118 samples in two registration-form-image databases. Some possible measures of efficiency are computed, and in each case the proposed approach performs better than traditional methods.

Natural language morphology integration in off-line Arabic optical text recognition

In this paper, we propose a new linguistic-based approach called the affixal approach for Arabic word and text image recognition. Most of the existing works in the field integrate the knowledge of the Arabic language in the recognition process in two ways: either in post-recognition using the language of dictionary (dictionary of words) to validate the word hypotheses suggested by the OCR or in the course of the recognition process (recognition directed by a lexicon) using a statistical model of the language (Hidden Markov Model or N-gram). The proposed approach uses the linguistic concepts of the vocabulary to direct and simplify the recognition process. The principal contribution of the proposed approach is to be able to categorize the word hypotheses in words that are either derived or not derived from roots and to characterize morphologically each word hypothesis in order to prepare the text hypotheses for later analyses (for example, syntactic analysis; to filter the sentence hypotheses).