Abstract

In this paper we develop a system for writer identification which involves four processing steps like preprocessing, segmentation, feature extraction and writer identification using neural network. In the preprocessing phase the handwritten text is subjected to slant removal process for segmentation and feature extraction. After this step the text image enters into the process of noise removal and gray level conversion. The preprocessed image is further segmented by using morphological watershed algorithm, where the text lines are segmented into single words and then into single letters. The segmented image is feature extracted by Daubechies’5/3 integer wavelet transform to reduce training complexity [1, 6]. This process is lossless and reversible [10], [14]. These extracted features are given as input to our neural network for writer identification process and a target image is selected for each training process in the 2-layer neural network. With the several trained output data obtained from different target help in text identification. It is a multilingual text analysis which provides simple and efficient text segmentation.

Authors

P. Mathivanan¹, B. Ganesamoorthy ², P. Maran³
Velammal Engineering College, India¹, Adhiparasakthi Engineering College, India², SSN College of Engineering, India³

Keywords

Slant Correction, Morphological Watershed Algorithm, Daubechies’5/3 Integer-to-Integer Wavelet Transform, Neural Network