Novel Technique for Layout and Handwritten Character Recoginization in OCR

Abstract

In the document image analysis document segmentation is very important step. Document segmentation is the process in which we segment the document which contains the heterogeneous data means data like printed text, handwritten text, graph etc. We do the document segmentation because our optical character recognition system is unable to recognize the whole document with multiple data type so before the recognition we have to apply the document segmentation so to define the each region correctly. We would be using document segmentation on the handwritten bills which contain the heterogeneous content thereby segmenting the text and non- text region and the text into printed text and handwritten text and then we classify the text region into printed text and handwritten text. Information energy approach has been used to segment the text lines into rows that can be embedded into the notepad and command window later which help to save the bill copy in e-format.