Design And Implementation Of Scanner Plugin And Scan Image Correction Algorithm Under Domestic Operating System

Next to keyboard and mouse,scanner has become one of the most important computer input devices,widely used in graphics and image processing,printing and advertising industry and other fields.As one of the important external computer equipment,in order to realize the independent control of its hardware and software,it is of great significance to develop the scanner backend under the domestic operating system.At the same time,in the actual scanning process,the images obtained by scanning media generally have the problems of image tilt and cropping difficulty.Through the design and implementation of the scanning image correction algorithm,in addition to ensuring that users can get the correct scanning images when using the scanner,it is also of positive significance to promote the study of the scanning image correction algorithm.The main work of this thesis is summarized as follows:(1)In order to develop the corresponding backend for the sheet-feed scanner developed by the team under the domestic operating system.In this paper,we choose to develop the scanner backend based on SANE standard and provide various functional interfaces to control the scanner for the application program.The communication between the backend program and the underlying device is realized based on the Libusb library.Finally,the SANE frontend under the domestic operating system was realized to control the feed-paper scanner developed by the team.(2)In order to realize tilt correction and background cropping of the scanned image,the edges of the scanning medium must be obtained.Sometimes,because the color of the scanning medium is too dark,it is close to the background color,the edge is not obvious.In this case,it is difficult to correct the image.Existing tilt correction algorithms and contour extraction algorithms are too complex to meet the speed requirements of sheet-feed scanner.In this paper,a more targeted edge feature point extraction method is proposed by finding some regularities in the background image data of scanned images.This paper first introduces the data characteristics and the reasons for the emergence of the characteristics in detail.At the same time,the related image correction algorithm is introduced,that is edge detection combined with line detection technology.Then,according to the data characteristics of the background image,the image correction algorithm is designed and implemented.Finally,the algorithm is tested and the results are analyzed.The scanner backend has met the basic requirements of the SANE front end for controlling the scanner.In the future,the functions of the backend can be expanded and modified according to the needs.And the image correction algorithm,for the scanning image with the data characteristics,whether it is the conventional scanning medium or the scanning medium with dark color and no obvious edges,the correction accuracy of the algorithm is very high,and the speed is also very fast.Therefore,this algorithm can effectively solve the problem of correction of scanned images by sheet-feed scanner.