i need line word and character segmentation matlab code
Posts: 14,118
Threads: 61
Joined: Oct 2014
The segmentation of the line, the word and the character are one of the critical phases of Optical Character Recognition (OCR). Due to the imperfection in segmentation, most of the recognition system produces a low recognition rate. Few works have been done for optical character recognition in the other Indian script however in case Manipuri language is almost insignificant.
The segmentation of a text-document into lines, words and characters, which is considered the crucial phase of preprocessing in Optical Character Recognition (OCR) is traditionally performed in uncompressed documents, although most documents in real life Are available in compressed form, for reasons such as transmission and storage efficiency. However, this implies that the compressed image must be decompressed, which indicates the additional computing resources. This limitation has motivated us to carry out investigations in the analysis of images of documents using compressed documents. In this article, we think of a new way of performing line-level, word, and character segmentation in compressed print-length document text documents. Extract the curve of the horizontal projection profile of the compressed file and using the local minimum points we do the segmentation of the line. However, tracing vertical information that leads to tracking character-words in a compressed run-length file is not very straightforward. Therefore, we propose a novel technique for simultaneously performing word and character segmentation, by skipping columns from each row into an intelligent sequence.