Design of Image Text Extracting Experiment Based on Stroke Filter
Liu Chun-pin · Research and Exploration in Laboratory · 2013
This paper proposed an experimental scheme of image text extracting method based on Stroke filter which consists of text location,text confirm and Stroke extraction three parts. The method takes full advantage of structure attribute of Strokes. With the help of the series Stroke response,it removes some noise by morphology method,then forms candidate text region,and text area is confirmed by using the heuristic knowledge. Finally color clustering method is used to divide text from background and obtain binary text image. Experimental results show that the method has better robustness,and suit extracting image text of different kinds. This design of image text extracting is helpful to train the undergraduate ability of solving the practical problem using learned related knowledge in image processing and analysis.