| The Chinese gravure printing industry has been restricted by high quality and high speed platemaking technology since its development,how to solve this key technology is an important issue that our country currently faces in this field.Lasers are widely used in this field due to their high processing efficiency and high focusing accuracy.Digital halftone technology is not only the core technology of the laser engraving machine's scanning engraving method,and it is also one of the important supporting technologies for laser plate making,the main solution is to solve the contradiction that the binary output device can only perform two operations and cannot directly output rich grayscale images.Therefore,the digital halftoning algorithm plays an important role in the reproduction of the subtle levels of the image on the laser plate surface and the overall efficiency of the plate making.At the same time,the image quality on the surface of the laser plate is also related to the quality of the laser-engraved cell.Therefore,it is also of great significance to use a machine vision-based quality discriminating method for gravure cells.The experimental equipment of this article include Shanghai Dianxiang SH-DXA gravure proofer,SCHEPERS laser engraving machine,Epson V300 scanner,MZDH0-850 C microscope ingest device,etc.and matlab2014 a simulation software,the experimental object is a copper plated gravure plate cylinder,a digital halftone method adapted to laser engraving of multiple grayscale images was studied and a gravure network quality detection system based on machine vision was established.The content as follows:1.Based on the traditional error diffusion algorithm,a gradient-based edge detection method is used to determine whether a pixel in the input image is a boundary point,then use the edge-enhanced quantizer to enhance the image edges,combining disordered ideas with error diffusion algorithms for halftoning.2.On the basis of Shiau-Fan algorithm error transfer,the edge of the image is extracted by the canny operator and the error filter coefficients are continuously optimized along the edge direction to achieve the edge enhancement effect of the halftone image.The experimental results show that the image obtained by the improved algorithm has good visual effects,less unwanted textures and more detailed edge details.Its PSNR value,WSNR value and UQI value were increased by 2.9-3.3 dB,2.5-3.2 dB,and 0.02-0.03,respectively.3.After the fast edge detection of the structured forest on the three dimensions of the original image,the result map is averaged,and the edge information of the image is extracted using an adaptive threshold.and we establish the texture partition model by combining the human visual characteristics with the local texture information sensitivity of the original image and combine it with the extracted image edge features.We use the SF error filter kernel to add the white noise random generator to halftone the multi-gray image.It can be seen that the proposed algorithm is closer to the blue noise characteristic by calculating the half-tone average power spectral density map of a uniform gray image with a gray level of 1/ 32,the PSNR value and MSSIM value of the halftone image obtained by the proposed algorithm are improved by 1.92-5.53 dB and 0.19-0.96 respectively compared to the other four methods.Experimental comparison of the algorithm using 40 standard images also validates the effectiveness and versatility of the proposed algorithm.4.After the grayscale linear transformation is performed on the gravure network image,the edge information is extracted using the roberts operator,calculate the ditch value and dark tone value of the cell and compare it with the standard value determined by the engraving process to judge whether the quality of the cell site is qualified.The evaluation results of the smoked proofs and the smoked images were processed in computer by different halftoning methods based on the evaluation index of the human visual system. |
PDF Full Text Request