A Study on Text Detection Techniques of Printed Documents

In the current era of technology advancements, the number of digital transmission libraries are growing terribly quickly, the necessity to expeditiously index, browse and retrieve transmission data is hyperbolic. In this paper, a technical analysis on textual data identification and extraction from image is given in which various approaches for text detection and extraction are discussed and an evaluation is done on the popularity of approach. Models and their techniques are compared with the help of taxonomy which aims to show how the work flow can be described by various techniques and a study of different classifiers used by different models is also given. Results of various models are shown in the form of recall and precision and we have found that 0.1% of the algorithms achieved more than 99% recall and precision and on the basis of it efficiency of model is found.