In 2003, the first image with the spam text inside was reported by Graham-Cumming. Later, this technique was utilized successfully by spammers, by sending image spam as MIME attachments instead of sending as simple image tags. The previous content filtering techniques based on text analysis of subject and body fields of email were ineffective to handle this new spam attack type. The first attempts made by researchers to detect such spam were based on Optical Character Recognition (OCR) methods. These methods tried to extract the spam texts/words from image spam and compare with existing spam text keyword database. This chapter provides the details of OCR methods, a literature review on spam filters based on OCR methods and their limitations.
Top4.1. Optical Character Recognition (Ocr): Introduction
Optical Character Recognition (OCR) is a pattern recognition technique which involves the process of converting the printed/typed or handwritten text (usually in the form of images) into machine-encoded text. Figure 1 shows the simple block diagram of OCR Reader.
In data entry applications like passport documents, invoices, bank statements, computerized receipts, this technique is widely adopted. This method of digitizing printed texts helps to edit electronically the converted text and hence in turn, offers benefits like ease of searching, storing data in compact form, ease of displaying data on-line etc.
4.1.1. History
The history of OCR can be traced all the way back to 1809, when reading devices for the blind or telegraph applications were developed. In 1914, Emanuel Goldberg developed a machine to convert printed characters into standard telegraph code. Concurrently, Edmund Fournier developed a handheld scanner called Optophone, which can produce tones corresponding to the specific letters/characters in the printed document (Herbert, 1982, Dalbe, 1914). In the late 1920s, Emanuel Goldberg developed an optical code recognition system for searching microfilm archives. In 1950s, US Department of Defense created GISMO, a device that could read Morse Code as well as words on a printed page, one character at a time. In 1974, Ray Kurzweil developed omni-font OCR reading machine for the blind, which could recognize text printed in virtually any font. This device utilized the CCD flatbed scanner and the text-to-speech synthesiser. Later commercial version of the OCR computer programs were launched in the market for commercial purposes like uploading legal paper and news documents onto online databases. In the early 1990's, it was used by libraries for historic newspaper digitization projects. An open source GUI frontend PrintToBraille tool (Rose, 2009). was developed by A. G. Ramakrishnan and his team at Medical intelligence and language engineering lab, Indian Institute of Science, to convert scanned images of printed books to Braille books. In the 2000s, WebOCR – an online OCR tool was made available in a cloud computing environment and in mobile applications. Currently, many commercial open source OCR systems are available that support other language writing systems.
4.1.2. Applications
OCR engines are used for the following applications prominently:
- 1.
Data Entry: It is widely used for data entry in business documents such as passport, check, invoice, bank statement, receipt, etc.
- 2.
Information Extraction: It has proved its worth in insurance sector by extracting important / key information from information documents.
- 3.
Business Card Information Extraction: Business sector is using it for creating the contact list of the card information.
- 4.
Textual Version of Printed Documents: Textual versions of the printed documents eg book scanning are quickly generated by this technique.
- 5.
Electronic Image: Formation of electronic images of searchable printed documents, like Google Books is carried out by this technique.
- 6.
Assistive Technology: It is used to make the assistive technology for blind and visually impaired users.