This paper is about neural approach for recognition of optical or visual characters. Previously the neural approach was implemented for optical characters but this paper suggests about simplified neural approach.
In the existing system classical approach for pattern recognition is not well suited as the characters can differ in size, shape and style from one person to other or can change within the course of time with the same person, even the visual characters can get spoiled due to noise. And also there are no rules for defining visual characters.
In the proposed system optical character recognition was suggested as it has advantages over the existing system, the new method implemented is adaptive with minimum errors and changes in patterns and the system introduced is a general system which is independent of size and aspect ratio and also user specific.
In general a software implements good documentation improves user friendliness and also maintainability whereas quality software results in user satisfaction where all the user requirements are met with no or less errors and the project all these features are well implemented.
The paper is about recognition of optical or visual characters and various approaches. In this paper we have dealt with the various approaches for recognition and its features and also introduced neural approach for recognition and compared its features to the existing system and concluded that the neural approach is the best approach which is simpler than other approaches.
Hardware required:
CPU Type : Intel Pentium IV
Display type : CGA, VGA, and SVGA.
RAM : 256 MB and above.
Software:
Front end : JAVA
Operating systems : windows 98, windows xp and higher versions.
Download Approach For Recognition Of Optical Or Visual Characters Project.