Abstract
Document Image Analysis (DIA) is the process that performs the overall interpretation of document images. DIA has many applications and some of the important applications are automatic data entry, reading aid for the blind, bank automation, post office automation etc.
At the beginning of the talk an overview of document image analysis will be discussed. Next, details about Optical Character Recognition (OCR) techniques of printed documents will be discussed. This includes (i) pre-processing like noise cleaning, skew detection and correction, text/graphics separation, lines, words and characters segmentation etc. (ii) feature extraction techniques (iii) classification techniques and (iv) post-processing.
A document page may contain several language scripts (e.g. because of the multi-lingual and multi-script behaviour an Indian document may contain two or more scripts in a single page). So, it is necessary to develop multi-script OCR systems for such documents. We shall also discuss about the multi-script OCR development technologies in this talk.
Next, we shall discuss about handwritten document recognition techniques. Application of handwriting recognition in postal automation and different steps for the development of postal system will be discussed. We shall also discuss about bank automation.
In the present age of computer, many printed documents are written in artistic way to draw people’s attention. In these artistic documents text lines of a single page may not be parallel to each other. These text lines may have different orientations or the text lines may be curved in shape. Because of multi-oriented behaviors of the characters, it is very difficult to develop OCR for such artistic documents. In this lecture, we shall also discuss about the recognition of artistic documents.
Finally, some DIA systems will be demonstrated. |
