Writer Recognition System: Study about the Simple and Hypride Source code for Write Biometric Recognition, based on Edge-Hinge Distribution. Writer Recognition System enables an individual to scribble something on a piece of paper and then convert it into text. In reality, there are numerous styles in which a character may be written. Furthermore, the combination can be done with these styles to generate more styles. If you know the basic styles in which a character can be written, then you would be able to recognize the characters formed by the combination of those basic styles. This logic stimulates the use of Genetic Algorithms for writing recognition. In order to prove the effectiveness of this algorithm, we made a pool of images of characters. And then, we converted them to graphs. The study showed that the graph of each character was intermixed with another to form styles intermediate between the styles of parent characters. Write recognition algorithm involved the matching of a graph generated from the unknown character image with the graphs generated by combining the characters of basic styles. The code for this method has been tested by IAM Handwriting Database. This algorithm is proved to be 98.44% accurate.
Tomas Langer, Student
The code on this project(Writer Recognition System) is very simple,smooth and easy to understand. Even though it is written by professionals it can be easily understood by a student who was just introduce to matlab programming. I added some of my own code to it for my semester project.
Katja Nacevski, Student
I came to this site because I saw free source code for Writer Recognition System .And as I went through the website I saw that they provides online matlab training. To this time, I have completed number of courses on matlab programming with them and I can see progress on my skills.