汉字激光标刻路径优化算法研究

发布时间：2018-03-26 08:18

本文选题：激光标刻　切入点：遗传算法　出处：《辽宁大学》2017年硕士论文

【摘要】：激光标刻技术随着各方面技术的快速发展和在工业、科研、国防等领域的广泛应用而不断更新发展,由于激光标刻系统是包含了机械、计算机、电子和光学等技术的一门综合技术,其中任何一门学科出现变革都会推动激光标刻技术进一步更新发展。现阶段,激光技术迅速发展,机械技术精度不断提高,计算机技术的发展更是掀起了一股时代的热潮。在我国的激光标刻应用中,汉字激光标刻最为常见,而且应用于工业领域的汉字轮廓激光标刻对标刻系统的速度有着相当高的要求,本文的主要工作便是以此为特定场景展开的。首先,从TrueType字库中提取汉字轮廓线。根据汉字在计算机内的代码得到汉字的轮廓线信息,包含在轮廓线上的控制点信息,将提取的信息存储到指定的数组中,再根据TrueType字库文件的特点以及包含的描述表将汉字轮廓线打印到显示器屏幕中。在汉字轮廓线提取过程中选取不同字体的汉字,并获得初始默认空行程路径。其次,针对特定的汉字轮廓激光标刻路径问题提出优化算法,优化算法主要以微种群遗传算法为基础,嵌入蚁群算法作为微种群遗传算法的适应度评估函数,运用改进的微种群遗传算法对标刻路径的空行程进行优化,在整个优化过程中,遗传算法中传统的交叉操作将无法解决汉字封闭轮廓线上控制点构成的环状索引在交叉方面出现的种群退化问题,这里将提出对环状索引进行变形,以解决交叉难题。最后,对不同字体的汉字轮廓线进行标刻路径优化实验,得到最终空行程优化百分比。实验中添加了对比实验,采用独立微种群遗传算法加蚁群算法作为对比实验,得出空行程优化百分比。最终对两个优化结果做出对比和分析。此外,对多种字体的字库做了优化实验与分析。本文以振镜扫描式标刻系统为研究系统平台,在Visual Studio 2008软件环境下,借助OpenCV视觉类库,使用C++语言完成汉字轮廓线的提取、空行程路径的优化以及优化结果的计算,最终输出优化结果。
[Abstract]:With the rapid development of laser marking technology and the extensive application in industry, scientific research, national defense and other fields, laser marking technology has been continuously updated, because the laser marking system includes machinery, computers, etc. A comprehensive technology of electronics and optics, in which any subject changes will promote the further development of laser marking technology. At this stage, laser technology is developing rapidly, and the precision of mechanical technology is constantly improving. The development of computer technology has set off an era of upsurge. In the application of laser marking in China, Chinese character laser marking is the most common. And the Chinese character contour laser marking applied in the industrial field has quite high requirements for the speed of the marking system. The main work of this paper is based on this for a specific scene. First, The contours of Chinese characters are extracted from the TrueType font. According to the code of Chinese characters in the computer, the contour information of Chinese characters is obtained, including the information of control points on the contour lines, and the extracted information is stored in a specified array. Then according to the characteristics of the TrueType font file and the description table, the contour line of Chinese character is printed to the display screen. In the process of extracting the outline line of Chinese character, the Chinese characters with different fonts are selected, and the initial default empty stroke path is obtained. Secondly, An optimization algorithm is proposed for the specific Chinese character profile laser marking path problem. The optimization algorithm is mainly based on the micropopulation genetic algorithm, and the ant colony algorithm is embedded as the fitness evaluation function of the micropopulation genetic algorithm. The improved micropopulation genetic algorithm is used to optimize the empty stroke of the mark path, and in the whole optimization process, The traditional crossover operation in genetic algorithm will not solve the problem of population degradation caused by the ring index formed by the control points on the closed contour line of Chinese characters. Here, we will deform the ring index in order to solve the problem of crossover. The optimization experiment of Chinese character contour with different fonts is carried out, and the percentage of the final empty stroke optimization is obtained. A comparative experiment is added in the experiment, and the independent micropopulation genetic algorithm and ant colony algorithm are used as the contrast experiment. The optimized percentage of empty stroke is obtained. Finally, the comparison and analysis of the two optimization results are made. In addition, the optimization experiment and analysis of the font library are made. In this paper, the vibrating mirror scanning marking system is used as the research platform. In the Visual Studio 2008 software environment, with the help of OpenCV visual class library, C language is used to complete the extraction of Chinese character contour, the optimization of empty stroke path and the calculation of optimization results, and the final output optimization results are obtained.
【学位授予单位】：辽宁大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN249;TP18

【参考文献】