Automation of Business Cards

Abstract

Business card is shared as hardcopy, the data present in business card will be highly useful if it is available in digital format. The task of manually entering the details of all business cards is laborious and time-consuming. Document image analysis is used in this paper for automating this process. This will be accomplished by performing OCR and then using the text to extract the Meta data. One more important component of business card is the logo of the organization. The text extraction OCR will be done using the Tesseract API. After conversion of the image to text, the data will be saved in the database. The raw data will be saved in the database, which will later be segregated and stored in the appropriate fields. It is generally ignored in the process of saving text information, in this paper it is extracted and stored in database. For logo detection various techniques like Gabor Filter, Harris edge detection technique, MSER, etc., are compared to determine the best technique for acquiring the most accurate logo extraction. Gabor filter gives the best result is used for the extracting logo and storing in database. Java language on NetBeans IDE platform which use the Spring MVC framework is used to implement this work.