The existing technology of color separation, especially the spot color separation, can no longer meet the requirements of prepress processing efficiency or printing quality. Concerning this situation, the Fuzzy C-Means ( FCM) clustering algorithm was put forward. The algorithm, based on the classification of pixels, carried fuzzy clustering on the grayscale of images in order to get image clustering center at first, and then put each pixel to the corresponding category according to the grayscale of each pixel and the maximum membership degree. The experimental result shows that image segmentation based on fuzzy rules is intuitive and easy to realize and has achieved a good segmentation effect.%现有印刷图像专色分色技术已经不能满足印前处理效率和印刷质量要求,针对这一现状,提出了一种模糊C-均值(FCM)聚类算法.该算法基于像素分类,它首先对图像的灰度级进行模糊聚类,得到图像的聚类中心,然后根据每个像素点的灰度级,依照最大隶属度原则将各个像素点归于相应的类别中.实验证明,采用FCM对印刷图像进行分割具有直观、易于实现的特点,实现了较好的分割效果.
展开▼