Handwriting Quality Assessment using Structural Features and Support Vector Machines

Handwriting plays a major role in written communication. Based on appealing handwriting students gather better scores and professionals become more successful. Timely feedback on the quality of handwriting helps to make course corrections. In this work, we report a system that takes in a scanned image of document to analyse & provide a quality indicator for the handwriting. The study performs a comparative analysis of various well-known classifiers (such as SVM, k-NN) using structural features (such as slant, inter-character spacing, character size). The features are obtained from binarized images as well as skeletonized images. It is observed that SVM classifier generates the maximum accuracy with the feature-sets.