目的 在高速印刷过程中，需要快速准确地对印刷品的质量进行检测，并给出数据评价，为此开发一种印刷网点增大的在线检测系统。方法 通过图像采集系统获得实际印刷的网点图像，基于机器视觉技术，利用Matlab高级程序语言实现网点图像的识别。结果 通过将标准网点图像与实际采集网点图像进行对比，利用有效的识别算法对3种形状印刷网点在50%处的增大量进行计算。结论 该识别算法适用性强，可应用于彩色网点印刷中多种形状网点增大量的在线识别。

The work aims to develop an online detection system for printing dot gain for the purpose of quickly and accurately detecting the print quality in the high speed printing and giving data evaluations. The printing dot images were obtained through the image acquisition system firstly. Then, the dot images were identified by means of the machine vision technology and Matlab advanced programming language. By comparing the standard dot images with the actual dot images, the effective recognition algorithm was used to calculate the gains of the three shapes of printing dots at 50%. The recognition algorithm has strong applicability and it can be applied in the online identification of gains of all shapes of dots in the color dot printing.