An Efficient Method to Extract Digital Text From Scanned Image Text

Abstract

Extraction of text from an image document is one of the challenges faced by nowadays. The paper focuses on the problem of text detection from scanned image documents for the improvements in its conventional techniques. The paper uses the Optical Character Recognition (OCR) technique in order to extract actual text presented in the input image. Other techniques such as Maximally Stable Extremal Region (MSER) to estimate the scales and orientation CC extraction which is used as an algorithm in order to enhance the extraction and retrieval process.