Deep Strip-Based Network with Cascade Learning for Scene Text Localization

Scene text detection is currently a popular research topic in the computer vision community. However, it is a challenging task due to the variations of texts and clutter backgrounds. In this paper, we propose a novel framework for scene text localization. Based on the region proposal network, a Stri...

Full description

Saved in:
Bibliographic Details
Main Authors: Dao Wu, Rui Wang, Pengwen Dai, Yueying Zhang, Xiaochun Cao
Format: Conference Proceeding
Language:English
Subjects:
Clustering algorithms
CNN
Hard Example Mining
Microsoft Windows
Proposals
Recurrent neural networks
RNN
Scene text detection
Strips
Task analysis
Training
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Scene text detection is currently a popular research topic in the computer vision community. However, it is a challenging task due to the variations of texts and clutter backgrounds. In this paper, we propose a novel framework for scene text localization. Based on the region proposal network, a Strip-based Text Detection Network (STDN) is developed with vertical anchor mechanism to predict the text/non-text strip-shaped proposals. Meanwhile, we incorporate the recurrent neural network layers in the proposed network to refine the predicted results. Specifically, hard example mining is performed to train the STDN with cascade learning, which has a remarkable improvement in precision. Besides, we exploit a clustering algorithm to generate anchor dimensions spontaneously without hand-picking, which is portable and time-saving. The text detection framework achieves the state-of-the-art performance on ICDAR2013 with 0.89 F-measure.
ISSN:2379-2140
DOI:10.1109/ICDAR.2017.140