Research On Halftone Screening Algorithm Based On Image Feature Best Matching

Halftone screening technology,generally,is used in the binarization processing of pre-press graphic information.In digital image processing,this technology is used to convert the continuous image signals obtained by photographing and scanning into binary dot image signals in the printing industry.It has been widely adopted in most printing methods such as offset printing,gravure printing,flexo printing,silk screen,digital printing,etc..With the development of computer information processing theory and the continuous improvement of related technical level,many screening algorithms have been proposed and applied,however,its screening method and dot characteristics are relatively simple,which do not fully restore the hierarchical detail features of continuous tone images,preserve complete image information,and correlate with the content features of images,resulting in the output halftone images being hard to reproduce the input continuous tone images with high fidelity,it's also affects and restricts the quality of printed matter to a great extent.In this paper,a halftone screening algorithm,which based on the best matching of image features and the the matching screening between the content features of continuous tone images and the dot features of halftone images,is designed.Firstly,based on Floyed-Steinberg's error diffusion algorithm,the weight coefficients are randomly scrambled in each diffusion,and the error diffusion weight coefficients with higher randomness are gained,which disrupts the iteration rules in each error diffusion algorithm and reduces the generation of interference fringes.Edge contour information,then,is extracted,which the image edge contour and texture feature information are enhanced;and extended to distinguish or extract the content features of the original continuous tone image,and adaptively match the mesh points with different sizes,shapes and angles according to the different contents and features of the image,so that it can restore the edge and texture features of the continuous tone image more realistically while satisfying the continuous tone of halftone reproduction,and make the halftone image achieve the best match between the content features and the original continuous tone image.Take different screening methods according to different characteristics,which makes the distribution of network points shows the trend of randomization that not only the tortoiseshell can be eliminated,but also the feature information of the image is highlight.The halftone image output by this technology has an excellent visual effect,make the texture part is presented more clearly,and the edge part better reproduces the gradation level of the continuous tone image.It can also reduce the influence of the interference of printing equipment and reduce the printing cost,which is more in line with people's needs,meets the development of the times and has a broader application prospect.