【作者】 曾强；

【导师】 戚晓芳； 马飞；

【作者基本信息】 东南大学 ， 工程硕士（专业学位）， 2017， 硕士

【摘要】 漏白是指在印刷过程中,往往会由于多个原本密接的色块位置不准或者计算精度的限制导致在色块与色块交接处露出白色间隙,或者因纸张经过传送拉力、热胀冷缩等外力干扰的原因而产生油墨间隙。漏白现象是印刷过程中难以避免的问题,容易造成印刷制品的图像轮廓走样、视觉体验差,严重影响产品的质量。本文提出了一种基于边界跟踪检测的陷印方案(修补漏白),并提出了两个关键的改进算法:基于八方向链码的边缘跟踪检测算法和基于扫描线种子标志位填充算法。跟原有的陷印方案相比,该方案在漏白判断的精确度、处理效率和陷印的成像效果上都得到了改善提高,在此基础上设计和开发了基于边缘跟踪检测的陷印系统,主要功能包括漏白检测、漏白填充、图像平滑等。具体而言,本文工作包括如下几个方面:(1)设计和实现了漏白检测模块。该模块主要负责预处理的基本功能,包括图像的输入、根据图像边缘检测算法判断图像是否需要做陷印处理和保存漏白区域的相关信息。边缘检测算法采用图像的八方向链码来定义边缘坐标和方向坐标,提高了对图像漏白判断的准确性和获取整体轮廓的效率性。(2)设计和实现了区域填充模块。该模块主要负责对漏白区域进行像素补偿,包括填充边缘的判断、填充像素的选定,并提出了基于扫描线的种子标志位填充算法对漏白区域进行像素填充。填充算法采用水平或者垂直扫描线的方法来填充像素区间段,避免了原有的种子填充算法需要大量递归的问题,节省了堆栈空间,提高了填充效率。(3)设计和实现了色彩方案模块。该模块通过匹配漏白区域信息及图像边缘信息制定填充像素值为区域填充模块提供了色彩支持,包括内外边缘色的计算、渐变填充色的计算以及色彩分量值的计算。(4)设计和实现了后期处理模块。该模块主要负责对陷印完成的图像进行后期的处理,包括设计高斯滤波器对漏白填充区域进行平滑滤波处理,降低边缘色彩的阶跃度和图像噪声干扰,然后再将处理完的图片输出或保存。本文方案对实际应用进行了可行性验证与测试。结果表明,本文方案可更准确地提取漏白区域的轮廓特征,提高了漏白判断的可靠性、漏白定位的准确性和漏白填充的高效性,解决传统陷印策略中存在的漏白边缘区域填充的准确度不高、修补漏白的色彩成像效果不佳的问题,使印刷产品的质量得到明显提高。更多还原

【Abstract】 In the printing process,there are two factors that result in white leakage.One factor is that theposition of multiple originally adjacent color lumps is not accurate or limited accuracy calculationresults in the white space at the junction of colors.The other is that after transmission of tension,thermal expansion and contraction of external force such as interference,white leakage appears in the paper.It is difficult to avoid white leakage phenomenon in printing process.And it easily yields the variant of image contour and poor visual experience of printed products,which will seriously affect products quality.A trapping scheme(repair white leakage)based on edge tracking detection and two key improved algorithms are proposed in this paper.The two improved algorithms are image edge detection algorithm based on eight direction chain code and sweep line boundary flag filling algorithm.Compared with the traditional trapping scheme,our scheme improves the accuracy of white leakage judgment,processing efficiency and the color imaging effect of trapping.Moreover,on the basis of the scheme,a trapping system based on edge tracking detection was designed and developed.This system implements functionalities for white leakage detection,filling white leakage,image smoothing,etc.In particular,our main work is summarized as follows:(1)Design and implement a module for white leakage detection.This module is mainly responsible for preprocessing basic functionalities,including image input,deciding whether the image is required to be trapped or not and saving white leakage information with our image edge detection algorithm.Our algorithm uses eight direction chain code to define edge coordinates and orientation coordinates,which improves the accuracy of the white leakage judgment and the efficiency of extracting an overall outline.(2)Design and implement a module for color scheme.This module formulates pixel values of leakage area by matching white leakage information and edge information.It provides color support for region filling module,including the calculation of internal and external edge color,the calculation of filling in gradient color and color component value calculation.(3)Design and implement a module for region filling.This module is mainly responsible for pixel compensation in the leakage area,including determination of filling edges and selection of filled pixels.Moreover,in this module a boundary flag filling algorithm is proposed to draw the entire white leakage area.The filling algorithm uses the method of horizontal or vertical scanning line to fill pixel intervals,avoiding the problem that original seed filling algorithm requires a time consuming recursive process.It also saves the storage space,thereby enhancing the filling efficiency.(4)Design and implement a module for post processing.The module is mainly responsible for the post processing after the completion of trapping,which includes Gaussian filtering in white leakage area,reducing color step degree of edge pixel and noise interference,as well as outputting or saving the processed image.Finally,the feasibility of our proposed scheme is verified and tested.Results show that it accurately extracts the contour features of white leakage area.In addition,it improves the reliability of white leakage judgment,the efficiency of white leakage filling and the accuracy of white leakage detection.It solves the low accuracy of white leakage edge area filling and poor color imaging effect problems in the traditional color trapping,and then improves the quantity of printing products significantly.更多还原