Comparative Study and Implementation of Supervised and Unsupervised Models for Recognizing Handwritten Kannada Characters
Authors:
- 1 May 2018
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 774-778
Abstract
The objective of this paper is to classify characters written in Kannada, a south Indian language. The characters are extracted from written documents, segmented and processed using various image processing techniques such as contrast normalization, denoising, thinning etc. using numpy and OpenCV. The dataset used is a combination of the Chars74k dataset and a custom-made dataset. Classifiers have been implemented based on K Nearest Neighbors, Support Vector Machines, Inception V3 and Convolutional Neural Networks using OpenCV and Keras and a comparison of their accuracies has been provided. The CNN classifier has been able to achieve an accuracy of 99.84% on the Chars74k dataset.Keywords
This publication has 8 references indexed in Scilit:
- Cyclical Learning Rates for Training Neural Networks2017
- Rethinking the Inception Architecture for Computer Vision2016
- Template matching method for Kannada Handwritten recognition based on correlation analysis2014
- Character segmentation algorithms for Kannada optical character recognition2008
- A simple and efficient optical character recognition system for basic symbols in printed Kannada textSādhanā, 2007
- A statistical approach to line segmentation in handwritten documents2007
- Learning Segmentation of Documents with Complex Scripts2006
- A fast parallel algorithm for thinning digital patternsCommunications of the ACM, 1984