An Efficient Algorithm for Skew-Correction of Document Image Based on Cyclostyle Matching

The skew-correction of scanned document image is a necessary step undergone before some processing. For the purpose of practical, robust, real-time of the algorithm, this paper presents a new skew-detection algorithm that based on the gray projection cyclostyle matching of Document Image. The algorithm, which is based on the analysis of gray scale projection curves of horizontal and vertical direction of the list images, compares with them if its not homology,then solves the problem of how the accurate skew angle can be found. The skew-adjustment algorithm can solve the problem of image distortion using a dual linear interpolation algorithm. In the end, the authors illustrate their experimental results, which show that the presented algorithm can solve the problem of skew-detection and skew-adjustment of the scanned document image. It is an important practical significance skew-correction algorithm for a large number of scanned document images.