Home // ACCSE 2017, The Second International Conference on Advances in Computation, Communications and Services // View article

Text Location Algorithm Based on Graph-Cut Model with Unary and Binary Features

Authors:
Fengqin Yu
Yaya Liu

Keywords: text location; graph cut model; unary text features; binary text features .

Abstract:
In this paper, we propose a location algorithm of the text regions in an image based on a graph cut model with unary and binary features. First, the most stable extremal regions are detected as candidate regions. Then, we define the energy function using unary and binary features of the candidate regions. The text classification is obtained by the optimal segmentation. Lastly, the final positioning is obtained using text aggregation. The simulation results show that the proposed algorithm is better than several classic algorithms in terms of precision rate and standard measurement and can precisely locate text regions in an image.

Pages: 23 to 25

Copyright: Copyright (c) IARIA, 2017

Publication date: June 25, 2017

Published in: conference

ISSN: 2519-8459

ISBN: 978-1-61208-570-8

Location: Venice, Italy

Dates: from June 25, 2017 to June 29, 2017