Text Extraction Systems for Printed Images: A Review
Neeru Saxena, Humera Parveen · SSRN Electronic Journal · 2019
Numerous researchers have extended text extraction schemes using Optical Character Recognition, Content Recognition, Feature Extraction, Wavelets and many other algorithmic approaches of Image processing. Picture components are considered for processing as characters, words and lines. Authors have focused on extremely challenging situations to deal with variation in text due to orientation, style, size and alignment. A comparative review is reported for efficient Text Detection and Recognition from Printed Images (TDE-PI). Design model is elaborated with its contribution, existing algorithms , various methodologies, related characteristics and percentage of accuracy along with the error detection. Systems have been implemented and proposed for further research. This paper reviews the algorithmic approaches, the studies of various researchers, implemented systems and provides a clear route for identification of texts from images. It further assists you in understanding performance evaluation factors and utilize the benchmark data for relevant research.