A feature extraction technique based on character geometry for character recognition

TL;DR

This paper introduces a geometry-based feature extraction method for character recognition that utilizes character contour skeletons to improve pattern recognition accuracy.

Contribution

It proposes a novel geometric feature extraction technique based on character contours and skeletons for segmentation-based word recognition systems.

Findings

Feature vectors effectively represent character contours

Neural networks trained on these features achieve accurate recognition

System benchmarks demonstrate competitive performance

Abstract

This paper describes a geometry based technique for feature extraction applicable to segmentation-based word recognition systems. The proposed system extracts the geometric features of the character contour. This features are based on the basic line types that forms the character skeleton. The system gives a feature vector as its output. The feature vectors so generated from a training set, were then used to train a pattern recognition engine based on Neural Networks so that the system can be benchmarked.