Text Area Extraction Method Based on Edge-Pixels Clustering

摘要

In this paper, we propose a novel text area extraction method based on edge-pixels clustering and boundaries extension for multilingual text extraction. We also present a text area binary approach which can reduce the number of binary images when the text color polarity was unknown for high efficient post- processing like character segmentation. The experimental result shows that the proposed algorithm is effective with integrality as high as more than 99%.