4. DESCRIPTION
BACKGROUND OF THE INVENTION
1. Field of invention
This present invention relates to the field of character recognition and character generation. Character generation is totally a new field as far as computers and the internet are concerned, though there have been a few related work as described below in the next section.
2. Prior Art
There has been a lot of work in the literature for character recognition and handwriting recognition. Several algorithms have been proposed towards recognition. But there 1ms been no work relating generation of handwriting using software. Some related works are as follows.
a. United States Patent: 6,640,007. Niie et al. Their works in generation of
handwriting use electronic writing implement and method of electronic
writing. They use electronic stamp and seal.
b. United States Patent: 6,694,045. Chung et al. Their work is on generation
and verification of digitized signature. They use several feature extraction
techniques and store a signature using such features.
SUMMARY OF THE INVENTION
No existing work store the entire character set. They do not generate entire handwriting. They do not store the characters in the basic way of images or font files. Getting handwritten text from typed text has never been proposed.
This invention looks at the above said non-existent works. This invention proposes a utility feature of generating handwritten text from typed text that uses character and handwriting recognition as a subtask. It also combines an innovative usage of generated handwriting as to provide to the users, the option of sending e-mails in their personal handwriting.
DETAILED DESCRIPTION OF THE INVENTION
The user will be asked to write a particular text which covers the entire character set. Cases like presence of characters in the beginning of a word, end of a word, beginning of a sentence will all be taken into consideration. The user may provide the written text either as an image or a specific implementation may directly extract characters while writing using any attached input device. The image may be obtained by scanning after writing on a paper or by using any other direct input methods like mouse on a drawing area on screen, usage of light pen and sensitive surface and convert into image.

Now, if the user's text is via scanned image, having got the text in image format we have to extract the various characters, how they are written in the beginning of the word, middle of the word and end of the word. Capital letters are also obtained. The training text that the user needs to write depends on the level of support one is providing. If one is planning to provide generation of cursive handwriting, several ways of writing a character has to be taken into consideration. For example, m after v and a, is differently written. They now have to be recorded in some way or other. One method is to create a font file in the name of the user and store all the characters. Another method is to create image files of individual characters and store it as the user's characters (stated as font file in the future discussion).
Now the font file has been built. This contains all the necessary characters that would be written by a particular user. A user will be allowed to use only his font file. This is done to avoid any kind of forgery. Now the user will be provided an editor to type and edit text which he wants to write. After completing the text, the user can apply his font file to the text. Now, the text would be converted into his handwriting. The character set can also cover several languages as per the user's wish as Unicode supports many international languages.
One of the usages of this handwriting generation is in e-mail. The mail service provider has to request for the scanned image of the specified text in the user's handwriting. The image will be uploaded onto the server by the user and the character set extraction will be done at the server. A font file will be generated in the user's mail-id. The user can log onto any system in the future, type a text and send mail by applying his handwriting font which is available at the server. Now, the mail can be sent as an image which would avoid altering of text by viruses over the Internet.