Feature Extraction Techniques for Text Recognition of Indian Scripts

Abstract

Several research works has been done by many researchers on Optical Character Recognition system. But various work has been done is on Greek, Chinese, English and Japanese characters. There has not been sufficient quality of work on text recognition of Indian languages like Bangla, Marathi, Malayalam, Telugu, Gujarati, Kannada, Gurumukhi and Oriya. The development of handwritten text recognition (HTR) is an interesting area in pattern recognition. In HTR, the set of features are very important in selecting the appropriate feature that produces little classification error. In this paper, we have presented a study on feature extraction and classification techniques used for character recognition of Indian scripts. In the fast moving world with the amazingly expanding technology, character recognitions play a ample role by providing more scope to perform research in OCR techniques.A considerable advancement in the work associated with the recognition of handwritten and printed text has been reported in last few years. From the last few decades offline handwritten text recognition has gained a lot of interest of researchers. It is well known that each individual people have some different writing style, so it is very difficult to identify or recognize the handwritten characters or numerals. Based on data gathering process a concise classification of recognition system has been discussed in this paper. Several feature extraction techniques & classifiers like, diagonal feature extraction, transition feature extraction, K-NN classifier (K-nearest neighbour) & SVM classifier (Support vector machine) are also illustrated in this paper. The methodology for word recognition has also been briefly explained in this paper.