Text Localization Algorithm Based on Visual Saliency and Edge Density

Peng ZHANG,Rongyi CUI
DOI: https://doi.org/10.3969/j.issn.1671-5896.2017.03.015
2017-01-01
Abstract:In order to solve the problem of low precision in traditiona l text localization algorithm, a robust text localization method based on visual saliency and edge density is proposed from the perspective of natural scene tex t properties.Firstly, spectral residual theory is used to extract the saliency region of the image, then search for the regions of large edge density in the sa liency regions and construct candidate connected regions, using a small amount o f prior information to filter out non text area.The experimental results on the standard data set show that this method can obtain 70% comprehensive detection rate compared to the method of text region detection using edge feature.
What problem does this paper attempt to address?