Recognition of Handwritten Devanagari Script : A Survey

With the advancement in the technology the jobs for the human kind is becoming more and more tedious with millions and trillions of documents waiting to be processed. Certainly, such a daunting tasks are next to impossible to be completed manually. Over a period, the things are getting automatically done with the robots or intelligent computer systems. Optical character recognition is one of such recent fields that is used in automatic processing of the images captured by the high quality cameras. It finds one of the most popular uses in automated toll plazas. This paper represents survey about various optical character recognition techniques that have been researched upon by various authors in the past. OCR was generally used to recognize printed or typewritten documents. Character recognition has two categories i.e. Online Handwritten character recognition is defined as Writing with a special pen on a pressure-sensitive digital tablet and Offline Handwritten character recognition is defined as recognition of character by camera and scanner. To recognize offline handwritten Hindi text is very difficult and challenging task due to different writing styles are written by different person. This paper presents the recognition of offline handwritten Hindi characters using digitization , pre-processing, segmentation, diagonal feature extraction , zoning feature extraction and classification that is commonly used in Verifying Process , Recognizing process like bank cheques, Signature Verification, Postcode Recognition, Passport readers, offline document recognition.