Curved Text Detection Methods for Scenic Images : A Review

The objective of this paper is to present a comprehensive exploration of deep learning based curved text detection methods. The main goal of text detection is to determine the position of text from input natural scenic image The position of the text area in the image is obtained using text detection and localization techniques. In this current era of deep learning, computer vision applications have evolved and taken new routes, such as smartphone translation technologies, supportive ways for the visually impaired, and so on. This paper is aimed at summarizing and analyzing the major changes and significant progresses of curved scene text detection and recognition in the deep learning era. This paper tends to highlight the Challenges arises in capturing and recognizing curved text from natural images. To identify publicly released scene image data sets as well as appropriate standard evaluation methods for scenic images.