Weka-Based Classification Techniques for Offline Handwritten Gurmukhi Character Recognition

In this paper, we deal with weka-based classification methods for offline handwritten Gurmukhi character recognition. This paper presents an experimental assessment of the effectiveness of various weka-based classifiers. Here, we have used two efficient feature extraction techniques, namely, parabola curve fitting based features, and power curve fitting based features. For recognition, we have used 18 different classifiers for our experiment. In this work, we have collected 3,500 samples of isolated offline handwritten Gurmukhi characters from 100 different writers. We have taken 60 % data as training data and 40 % data as testing data. This paper presents a novel framework for offline handwritten Gurmukhi character recognition using weka classification methods and provides innovative benchmark for future research. We have achieved a maximum recognition accuracy of about 82.92 % with parabola curve fitting based features and the multilayer perceptron model classifier. In this work, we have used C programming language and weka classification software tool. At this point, we have also reported comparative study weka classification methods for offline handwritten Gurmukhi character recognition.