Research On Skew Correction And Denoise For Document Image

As the widely use of OA system, the carrier of document is quickly replaced by magnetic disk, magnetic tape and CD. The document image system turn the bills, notes, files and documents into digital images, keep and index them into the CD or magnetic disk by the way of document digitization. In this way, the document image system realizes the digital storage, administration and maintenance of the document and provides the network transfer and utilization. The document image system is used in the area of bank, insurance, government, communication, hospital and education widely. So, the research of document system not only is a meaningful question but also has a good grow foreground.The result of the digital images processing forms the final digital storage. So the quality of the image processing determines the quality of the digital document. The problems that the document image system needs to deal with can be divided into two aspects: the one is the optimization of the images, as noise elimination, skew correction; the other one is the specific requirement, as the digital watermark. This article deals with the two problem of image optimization: skew correction and noise elimination. This article provide a text-row-line based method to computer the skew angle of the document images and a text-area based method to center the text area. This article also provides a block based method to wipe the noise points and a neighbor-area based method to wipe the points between the words. At the end, the article gives a prototype system.