Text Extraction from Indian and Non-Indian Natural Scene Images: A Review

Natural scenic images usually contains textual data which may provide valuable information about the scene. To make this textual data useful, processing of scenic images involves various phases such as detection, localization, segmentation and recognition of text. First phase i.e. extraction of textual data plays an important role in further processing. Text extraction becomes difficult owing to variation in text font, size, skewness and noise in the captured image and is a challenge for the researchers. Over the years a lot of research has been dedicated to overcome these challenges and, this work presents extraction methods used for text in different Indian and non-Indian scripts. This article will provide academicians and practitioners with state of the art and future directions in this phase.