Text Localization in Natural Scene Images Based on Conditional Random Field

This paper proposes a novel hybrid method to robustly and accurately localize texts in natural scene images. A text region detector is designed to generate a text confidence map, based on which text components can be segmented by local binarization approach. A conditional random field (CRF) model, considering the unary component property as well as binary neighboring component relationship, is then presented to label components as "text" or "non-text". Last, text components are grouped into text lines with an energy minimization approach. Experimental results show that the proposed method gives promising performance comparing with the existing methods on ICDAR 2003 competition dataset.

1. Introduction

With the widely use of digital image capturing devices, text localization/detection, as a key part of the image text information extraction system [3], has been studied intensively. Though many efforts have been devoted to, it remains a challenge due to variations of texts' size, font, array orientation and degraded images with the cluttered background and noises [5]. The existing methods can be categorized into two classes: region-based methods and connected component (CC)-based ones.