Recovery of drawing order in handwritten digit images

This paper presents recovery of drawing order that converts offline handwritten text to their online handwritten format. The offline text images are preprocessed using stages as size normalization, noise removal and thinning of text. We have proposed a traversal algorithm to recover drawing order of digit images that convert offline image to online handwriting format and find trajectory direction using chain code features. The chain code features are extracted from the recovered trajectories and support vector machine has used as recognition technique to recognize text images. Our approach has been implemented with MNIST database and we have achieved an overall error rate as 2.61%.