Analyses of different methods of writing using SVM classifier

A person's handwriting reveals his personality and emotion, feeling like fear, happiness, honest or depression. The computer may find it difficult to detect a human motion because it is unaware of human emotion. To predict human emotions, various methods like facial recognition, eye tracking and analysis of handwriting are used. The scientific name of handwriting analysis is graphology; each person writing style is unique when compared to others. Analyzing each person handwriting in the offline method is difficult, costly and takes time. So, this paper is aimed at analyzing the different types of handwriting and extraction of various features. The features extracted will be trained for further classification. In this paper, a machine learning technique support vector machine is being implemented and discussed. The average time taken for writing a sentence was 17 seconds. This proposed system has accuracy varying from 92.33% to 62.5% while predicting the signal.