An Intelligent Approach for Recognition and Deletion of Strike-Out Text in Kannada Handwritten Scripts

Abstract

Handwritten document image processing is more demanding and  specific area in machine  and pattern recognition field. This research work presents a new approach to  identify strike –outs in Kannada handwritten document images. To address these issues, paper  proposes a hybrid model combining  the  feature based classifier  and graph-based techniques. The dataset is categorized into two classes: striked-out and non striked-out, and a Support Vector Machine (SVM) classifier is used for feature extraction in pattern classification. In the graph-based approach, we employ the shortest path algorithm to analyse strokes and also  addressed the complexities of zigzag or wavy style strike-outs.