Deep Scene Text Detection with Connected Component Proposals
arXiv:1708.05133
Abstract
A growing demand for natural-scene text detection has been witnessed by the computer vision community since text information plays a significant role in scene understanding and image indexing. Deep neural networks are being used due to their strong capabilities of pixel-wise classification or word localization, similar to being used in common vision problems. In this paper, we present a novel two-task network with integrating bottom and top cues. The first task aims to predict a pixel-by-pixel labeling and based on which, word proposals are generated with a canonical connected component analysis. The second task aims to output a bundle of character candidates used later to verify the word proposals. The two sub-networks share base convolutional features and moreover, we present a new loss to strengthen the interaction between them. We evaluate the proposed network on public benchmark datasets and show it can detect arbitrary-orientation scene text with a finer output boundary. In ICDAR 2013 text localization task, we achieve the state-of-the-art performance with an F-score of 0.919 and a much better recall of 0.915.
10 pages, 5 figures
References in corpus (8)
- Very Deep Convolutional Networks for Large-Scale Image Recognition
- Caffe: Convolutional Architecture for Fast Feature Embedding
- ParseNet: Looking Wider to See Better
- TextBoxes: A Fast Text Detector with a Single Deep Neural Network
- DeepText: A Unified Framework for Text Proposal Generation and Text Detection in Natural Images
- Multi-Oriented Text Detection with Fully Convolutional Networks
- Cascaded Segmentation-Detection Networks for Word-Level Text Spotting
- StuffNet: Using 'Stuff' to Improve Object Detection