Structure-Based Partition and Grouping For Text String Detection In Natural Scenes

Computer Engineering

Text in natural image perform excellent task for many image based applications used in computer vision like scene understanding, content-based image retrieval, assistive navigation, and automatic geocoding.With these applications ,now a days text detection and recognition became challenging and trendy task. However, locating text from a complex background with multiple colors is a challenging task. Here, a new framework is explored to detect text strings with arbitrary orientations in complex natural scene images. The proposed framework of text string detection consists of two steps: 1) image partition to find text character candidates based on local gradient features and color uniformity of character components and 2) character candidate grouping to detect text strings based on joint structural features of text characters in each text string such as character size differences, distances between neighboring characters, and character alignment. We propose two algorithms of text string detection: 1) adjacent character grouping method and 2) text line grouping method. With adjacent character grouping sibling groups of each character candidate get calculated and then merges the intersecting sibling groups into text string. The text line grouping method used to perform transformation to fit text line among the centroids of text candidates. The fitted line describes orientation of potential text string. The detected text string is presented by a rectangle region covering all characters whose centroids are cascaded in its text line.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.