基于Radon变换的文本图像倾斜校正

Skew Correction of Document Image Based on Radon Transformation

归纳现有的多种倾斜字符校正方法,提出一种基于Radon变换用于一般字符校正的方法。该方法对文本图像做简单的数学形态学处理,利用Radon变换求出文本倾斜的角度。实验结果表明:与传统方法相比,本算法减少了计算量,提高了校正效率,且具有较强的适用性和鲁棒性。

In order to improve the efficiency of the correction of skew document images, this paper proposes a new technique for skew characters correction. This algorithm processes the document images mathematically and morphologically, and finds out the skew angle of the text lines in the Radon transformation. The outcomes show that the new method significantly reduces the computation time, improves the efficiency of correction, and has proved to be applicable and robust.