印刷网点微观图像阈值分割算法研究

目的 通过阈值处理方法， 准确获取网点微观图像的特征参数， 将其与仪器测量值相结合， 综合评价印刷品复制质量。方法 提出一种基于高斯函数模型拟合网点图像灰度直方图数据的阈值分割算法， 寻找网点类图像最佳分割阈值， 对图像进行二值化处理， 得到准确的网点参数。结果 得到的印刷品网点面积率在全阶调范围内更接近于测量值， 分割效果明显优于传统的阈值分割算法。结论 提出的高斯拟合阈值分割算法更有利于提取网点类图像的微观参数， 精度高， 稳定性好，为获取准确的网点图像微观参数提供了理论与实践参考。

Threshold Segmentation Algorithm for Printing Dot Microscopic Image

In this paper, by the method of threshold segmentation, the characteristic parameters of dot microscopic image were accurately acquired which were used in combination with the instrument measured value to comprehensively evaluate the quality of printing. This experiment presented a threshold segmentation algorithm based on Gaussian function model to fit the gray histogram data of dot microscopic image to find the optimal threshold of the dot image and use this threshold to do binarization processing of image and getting accurate dot parameters. The obtained printing dot area rate by using this method was closer to the measurements throughout the tone range, the segmentation effect was better than the traditional threshold segmentation algorithm. In conclusion, Gaussian fitting threshold segmentation algorithm proposed in this paper is more beneficial to extract the microscopic parameters of the dot image, with high extraction precision and good stability, which provides a theoretical and practical reference for obtaining accurate dot image microscopic parameters.