作者: Yichao Ma , Chunheng Wang , Ruwei Dai
DOI: 10.1007/11940098_21
关键词:
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 in document images. According method, the image binarized first. Next, curved text-line features are extracted. Thirdly, locally optimized text curves detected using graph model. Finally, technique texture warping applied image. Experimental results show that restored by our can achieve good perception and results.