基于FCM和离散正则化的多目标图像分割.pdfVIP

第27 卷 第 1 期 计算机辅助设计与图形学学报 Vol. 27 No.1 2015 年 1 月 Journal of Computer-Aided Design Computer Graphics Jan. 2015 基于FCM 和离散正则化的多目标图像分割 1) 1)* 2) 1) 刘晓妮 , 卢奕南 , 雷玲 , 瞿子淇 1)(吉林大学计算机科学与技术学院 长春 130012)) 2) (华为技术有限公司质量部 深圳 518000) (liu_xiaoni@) 摘 要: 针对机器视觉中的多目标图像分割问题, 提出一种适用于多目标物体的图像分割算法. 首先对图像进行图 像增强预处理; 然后采用基于直方图的模糊 C 均值聚类算法完成分类任务, 实现图像的初分割, 将分类后的像素作 为种子集; 最后利用离散正则化的半监督方法得到自动修正分类结果. 实验结果表明, 与已有的多目标分割算法相 比, 该算法分割结果更加精确. 关键词：图像分割; 模糊C 均值聚类; 离散正则化 中图法分类号：TP391.41 A Multi-Objective Image Segmentation Method Based on FCM and Discrete Regularization Liu Xiaoni1), Lu Yinan1)*, Lei Ling2), and Qu Ziqi1) 1) (College of Computer Science and Technology, Jilin University, Changchun 130012) 2) (The Quality Department Huawei Technologies Co. Ltd, Shenzhen 518000) Abstract: To address the problem of multi-objective image segmentation in computer vision, a multi-objec- tive image segmentation method based on fuzzy C-Means and discrete regularization is proposed in this pa- per. First, the method preprocesses an input image with image enhancement. Secondly, the FCM clustering algorithm based on histogram is used to classify the pixels in the images into the different categories and realize the initial segmentations. Finally a discrete regularization algorithm as a semi-supervised method re- vises the classified results. The experiments demonstrate the superior performance of the proposed method in terms of segmentation accuracy. Key words: image segmentation; fuzzy C-means clustering; discrete regularization 图像分割是机器视觉中的一个基础问题, 同 分类与 Watershed 相结合方法分割多细胞图像. 时也是图像处理、图像分析、模式识别与跟踪的重 Grady[3]