Classification of Brahmi script characters using HOG features and multiclass error-correcting output codes (ECOC) model containing SVM binary learners

A lot of research has been done in the field of character recognition, but very less work has been reported related to Indian ancient scripts. Brahmi is considered as oldest script of India and is parent of all modern-day Indic scripts. In the process of optical character recognition, feature selection and classification are very of the important and crucial stages. Selection of appropriate, unique and simple features helps in better recognition. In this paper, authors have presented a system to recognize the Brahmi script characters using HOG features and SVM classifier with ECOC model. The presented system have achieved 92.09% of accuracy in recognition of Brahmi script characters.