QQ客服 官方微博 反馈留言

题名	图象处理在激光照排中的应用
作者	任群力
学位类别	硕士
答辩日期	1993-06-30
授予单位	中国科学院研究生院
导师	杨章运
中文摘要	由于激光照排中图形图象处理的复杂性，目前出版物中各类插图的排印效果与排印自动化程度还远不及对文字的处理。化学、化工学科出版物中有许多相图、化工过程流程图、化工装置设备图等图形图象，其排印效果的优劣是决定出版物质量的关键。本论文基于图形图象处理和识别理论，提出了针对激光照排中一类线图形的自动化处理方法。通常激光照排系统通过扫描仪完成插图的输入。由于扫描仪的机械和光电特性以及原图质量问题，使得输入后的插图产生噪声。特别对科技文章中以线条为主的插图，更容易出现噪声点、线中断、线扭曲、文字模糊、版面倾斜等现象，轻者影响版面的美观，重者导致版面失真。因此运用图象处理的方法，对扫描图进行有效的处理，解决输入图的质量问题是提高插图排印质量的关键。本文以二值图象基本算法为基础设计修图系统。建立以光栅数据为基础的预处理系统，完成噪声的消除，坐标校正，并通过文字过滤，实现字符与线条图形的分离，为下一步处理做好数据准备。实现了线图形的细化，图形骨的提取，从而提取线的几何特征，完成基于光栅数据的线段短枝消除和短缺弥合等处理。实现对图象光栅数据的矢量化，运用数学方法处理线状图形，采用直线拟合算法消除线段波动，实现断口连接等处理。实现对滤去文字的再生，利用标准字库中字模按比例放大、缩小的方法完成插图中文字的规格化处理。根据论文提出的一些基本方法，建立了初步的程序系统，该系统采用标准图象文件格式（PCX）作为软件输入，因此与其他标准化软件相兼容。另外，系统在设计中采用了最优组合的概念，从而使系统开销大为降低。通过对各类图形的处理结果表明，论文中提出的方法具有实际诮用前景。
英文摘要	The computerized laser typesetting has brought great progress in publication. To be with Chinese character processing technology, new graphs (images) processing technique is very important for the purpose of getting graphs on the typeface with higher quality. For the complexity of graphs's processing in typesetting, automatic processing of graphs is not as good as the text (words) processing so far. To improve the visualized effects of graphs in typesetting, new approaches are proposed based on the general theory of graph and image processing. Since the mechanic and photoelectric limitation of the scanner and the quality of original graph, noises are inevitably added to the scanned graph. In scientific journals, it often happens unexpected faults, such as the abnormal interception, twist of the curves, vague of texts, and even the oblique of whole graph. It is necessary to reconfigurate the graph so as to get an improved form of the graph on the typeface. To realize the reconfiguration of a specific graph, two stages are proposed in the thesis based on the fundamental algorithms for processing of two-dimensional graph. At the first stage, noises are detected and eliminated, coordinates are rescaled and text is also filtered out for further processing the curve on the graph. To perfect the graph, the second stage is thinning the curves with specified widths. According to analysis of the geometric features of the curve, through the extraction of their trunks, short fins and missing (optical grading) points can be found and erased, at the same time, the missing points can be filled. Mathematical methods have been applied in the procedure of graph reconfigurations to maintain the effect and efficiency of graph processing. To establish the vectors from the optical grading points of the graph, curve vibrant on the graph can be eliminated and interception can be patched. After the graph has been reconfigurated, text (words) on the graph filtered early is regenerated at their original positions, by utilizing the character module in a standard library. The programming system based on the approaches proposed in the thesis is compatible to the standards now commonly adopted in processing of graph. The hierarchical structure of the system optimizes the system execution and minimizes the expense of memory. Results of tests to graphs show the methods proposed in this thesis applied in typesetting are feasible and practical
语种	中文
公开日期	2013-10-31
页码	70
内容类型	学位论文
源URL	[http://ir.ipe.ac.cn/handle/122111/4657]
专题	过程工程研究所_研究所（批量导入）
推荐引用方式 GB/T 7714	任群力. 图象处理在激光照排中的应用[D]. 中国科学院研究生院. 1993.

个性服务

查看访问统计

相关权益政策

暂无数据

收藏/分享

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。