Abstract: In order to solve segment text accurately and robustly from a complex background, this paper proposed a new algorithm for text segmentation in images by using a stroke filter. First, we described the stroke filter briefly based on local region analysis. Second, the determination of text color polarity and local region growing procedures were performed based on the response of the stroke filter. Finally, the feedback procedure by the recognition score from an optical character recognition (OCR) module was used to improve the performance of text segmentation. The proposed algorithm was compared with other algorithms. The experimental results demonstrate that the proposed algorithm obtaines satisfactory results.

Key words: Text segmentation, Complex background, Optical character recognition(OCR) , Stroke filter