Correcting Bound Document Images Based on Automatic and Robust Curved Text Lines Estimation

摘要

Geometric distortion often occurs when taking images of bound documents. This phenomenon greatly impairs recognition accuracy. In this paper, a new one-image based method is proposed to correct geometric distortion in bound document images. According to this method, the document image is binarized first. Next, curved text-line features are extracted. Thirdly, locally optimized text curves are detected using a graph model. Finally, the technique of texture warping is applied to correct the image. Experimental results show that images restored by our proposed method can achieve good perception and recognition results.