[1]刘洋,周宁宁.基于SLIC的图像分割方法研究[J].计算机技术与发展,2019,29(01):75-79.[doi:10. 3969 / j. issn. 1673-629X. 2019. 01. 016]
　LIU Yang,ZHOU Ning-ning.Research on Image Segmentation Method Based on SLIC[J].,2019,29(01):75-79.[doi:10. 3969 / j. issn. 1673-629X. 2019. 01. 016]
点击复制

基于SLIC的图像分割方法研究()
分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
29
期数:
2019年01期
页码:
75-79
栏目:
智能、算法、系统工程
出版日期:
2019-01-10

文章信息/Info

Title:
Research on Image Segmentation Method Based on SLIC
文章编号:
1673-629X(2019)01-0075-05
作者:
刘洋;  周宁宁
南京邮电大学 计算机学院,江苏 南京,210046
Author(s):
LIU Yang; ZHOU Ning-ning
School of Computer Science,Nanjing University of Posts and Telecommunications,Nanjing 210046,China
关键词:
GrabCut;  图像分割;  超像素;  SLIC;  K-means
Keywords:
GrabCut; image segmentation; superpixel; SLIC; K-means
分类号:
TP301
DOI:
10. 3969 / j. issn. 1673-629X. 2019. 01. 016
文献标志码:
A
摘要:
图像分割是将图像分割成若干个特定的、独特的区域,并提取出目标的技术和过程.它是计算机视觉和图像处理中重要的研究内容之一.近年来,以GrabCut为代表的基于图论的图像分割方法成为了研究热点.基于图论的图像分割是一种将图像边缘信息和纹理信息相结合的能量最小化方法.GrabCut算法是一种优秀的基于图论的交互式图像分割算法.该算法具有简单的交互性和良好的分割效果,但也存在分割速度较慢、算法时间复杂度过高以及对背景和前景相似的图片分割不理想等缺陷.超像素块提取是指将图像分割成一组集合,集合的每个超像素块里的像素点都有相似的特征,各个超像素块之间有着明显的区别.超像素有许多便捷的属性,并被广泛地用作基础预处理阶段,以减少各种计算机视觉任务的计算.针对GrabCut图像分割方法时间复杂度过高以及对背景和前景相似的图片处理质量不好的问题,提出了基于超像素的GrabCut图像分割方法.
Abstract:
Image segmentation is the technology and process of dividing the image into several specific and unique regions and extractingthe target,which is one of the important research contents in computer vision and image processing. In recent years,image segmentationbased on graph theory,represented by GrabCut,has become a research hotspot. Image segmentation based on graph theory is an energyminimization method that combines image edge information and texture information. The GrabCut is an excellent interactive image segmentation algorithm based on graph theory. The algorithm has simple interactivity and great segmentation effect,but there are some defects such as slow segmentation,high algorithm time complexity and poor segmentation of similar background and foreground images. Superpixel block extraction refers to the segmentation of the image into a set of collections,where each pixel in each superpixel block has similar characteristics,and there are obvious differences between the superpixel blocks. Superpixels have many convenient properties andare widely used as a basic pre-processing stage to reduce the computation of various computer vision tasks. In view of the high timecomplexity of GrabCut image segmentation method and the poor quality of image processing with similar background and foreground,wepropose a method of GrabCut image segmentation based on hyperpixel.

相似文献/References:

[1]李嘉刚 李小宁 石杰 庄敏 陈戈.GrabCut在人体序列切片图像分割中的应用[J].计算机技术与发展,2011,(12):246.
　LI Jia-gang,LI Xiao-ning,SHI Jie,et al.Application of GrabCut in Human Serially Sectioned Image Segmentation[J].,2011,(01):246.
[2]马冬梅,武永娟,火元莲,等.一种改进的非局部极值FCM图像分割算法[J].计算机技术与发展,2018,28(09):20.[doi:10.3969/j.issn.1673-629X.2018.09.005]
　MA Dong-mei,WU Yong-juan,HUO Yuan-lian,et al.An Improved Non-local Extreme FCM Image Segmentation Algorithm[J].,2018,28(01):20.[doi:10.3969/j.issn.1673-629X.2018.09.005]
[3]李仔麒,马慧彬,李殿奎,等.改进区域生长法的肝部CT图像ROI提取[J].计算机技术与发展,2019,29(01):150.[doi:10. 3969 / j. issn. 1673-629X. 2019. 01. 031]
　LI Zi-qi,MA Hui-bin,LI Dian-kui,et al.ROI Extraction of Hepatic CT Images with ImprovedRegional Growth[J].,2019,29(01):150.[doi:10. 3969 / j. issn. 1673-629X. 2019. 01. 031]

更新日期/Last Update: 2019-01-10