kannada character segmentation code in opencv

kannada character segmentation code in opencv

Abstract

Data processing and management is common now a days. In this paper, automatic processing of forms written in Kannada language is considered. A suitable pre-processing technique is presented for extracting handwritten characters. Principal Component Analysis (PCA) and Histogram of oriented Gradients (HoG) are used for feature extraction. These features are fed to multilayer feed forward back propagation neural network for classification. Only 57 characters are used for recognition. Performances of two features are compared for different number of classes. HoG is found to have better recognition accuracy than PCA as number of classes increased. This is implemented in Visual Studio 2010 using OpenCV library. Keywords—Back Propagation Neural Network, Form Processing, Histogram of
Gradients, Kannada Script, Principal Component Analysis.