HANDWRITTEN NUMERICAL CHARACTER RECOGNITION

Handwritten Numerical Character Recognition (HNCR) is the process of interpreting handwritten digits by machines. There are several techniques in order to detect the handwritten digits. In this paper, it is proposed to the use of Histogram Orientated Gradient (HOG) feature extraction technique and Support Vector Machine (SVM) to detect the handwritten characters. HOG is a very efficient and stable feature in recognition system. Moreover, linear SVM has been employed as classifier which classify the handwritten characters with the help of Modified National Institute of Standard and Technology (MNIST) dataset with has a sample of 70,000. The combination of SVM and HOG is very efficient with MNIST dataset number classification. Moreover, the system is also test on 2 different single board computers to differentiate the performance of the 2 different systems. The primary scope of the project is to recognize and tabulate Student Exam Identity Number and Handwritten Marks.