基于Mojette变换的倾斜文本图像方向校正-《计算机技术与发展》

文章信息/Info

Author(s):: SHA Zhou-yang ; ZHANG Cui; [1]Department of Computer Science and Technology, Civil Aviation University of China[2]Department of Information Science and Engineering, Ocean University of China

Keywords:: Mojette transfrom ; tilt angle of the image; direction correction

摘要:: 文本文件的扫描图像在现实中几乎无处不在，文中将Mojette变换引入扫描支本图像使用前的方向校正，目的是为方便人们对扫描文本图像的使用。霍夫变换和矩形框调整算法是目前最常用的方向校正算法，但这些算法存在计算复杂度高和适应度差的缺点。文中方法是利用文字成行的特点，通过计算文字走向得到图像倾斜角，进而对图像进行校正。选择不同角度的倾斜图片进行了实验，均检测出文字走向，并成功校正图像。实验结果表明，该算法计算量小，适应度强，比较适合文本图像的方向校正

Abstract:: The manned image of text file has important applications in reality, the text direction correction before dealing the scanned image is essential to correctly identify the word. Hough transform and Rectangle adjustment algorithm are the mean algorithms in direction correction, but these algorithms exist the weak point such as high computational complexity and poor adaptability. Using the feature of the text trip arranged,it introduces the Mojette transform to solve the text direction correct problem,by computing the tilt angle of the image to get the text trend, the results demonstrate that the algorithm has small amount of calculation and strong fitness ,is more suitable for di- rection correction of the text image