Robust interactive image segmentation via graph-based manifold ranking
Robust interactive image segmentation via graph-based manifold ranking
复制标题
通过基于图的流形排序进行鲁棒的交互式图像分割
DOI：
10.1007/s41095-015-0024-2
复制
发表时间：
2015-09
影响因子：
6.9
通讯作者：
Enhua Wu
中科院分区：
文献类型：
--
作者：
Hong Li;Wen Wu;Enhua Wu
交互式图像分割的目的是在给定前景和背景标记的情况下，将图像像素分为前景和背景两类。在本文中，我们提出了一种新的交互式图像分割框架，它建立在基于图的流形排序模型的基础上，该模型是一种基于图的半监督学习技术，可以学习关于输入数据所揭示的内在结构的非常平滑的函数。通过克服传统模型中构建图的两个核心问题：图的结构和图的边权重，改进了最终的分割结果。用户提供的涂鸦被视为必须链接和不得链接的约束。然后，通过将这些约束和扩展的相邻空间关系集成到图结构建模中，将图建模为近似k-正则稀疏图。针对以往模型通常采用统一的核参数的不足，提出了内容和标签驱动的局部自适应核参数。在图的构建之后，提出了一种新的三阶段分割策略来得到最终的分割结果。由于我们构建的图的稀疏性和扩展的邻域关系以及超像素的使用，我们的模型可以提供几乎实时的、用户涂鸦不敏感的分割，这是交互式图像分割的两个核心需求。最后，我们的框架很容易扩展到多标签分割，对于一些不太复杂的场景，它甚至可以通过单线交互得到分割后的对象。实验结果和与其他先进方法的比较表明，该框架能够高效准确地从背景中提取前景目标。
Interactive image segmentation aims at classifying the image pixels into foreground and background classes given some foreground and background markers. In this paper, we propose a novel framework for interactive image segmentation that builds upon graph-based manifold ranking model, a graph-based semi-supervised learning technique which can learn very smooth functions with respect to the intrinsic structure revealed by the input data. The final segmentation results are improved by overcoming two core problems of graph construction in traditional models: graph structure and graph edge weights. The user provided scribbles are treated as the must-link and must-not-link constraints. Then we model the graph as an approximatively k-regular sparse graph by integrating these constraints and our extended neighboring spatial relationships into graph structure modeling. The content and labels driven locally adaptive kernel parameter is proposed to tackle the insufficiency of previous models which usually employ a unified kernel parameter. After the graph construction, a novel three-stage strategy is proposed to get the final segmentation results. Due to the sparsity and extended neighboring relationships of our constructed graph and usage of superpixels, our model can provide nearly real-time, user scribble insensitive segmentations which are two core demands in interactive image segmentation. Last but not least, our framework is very easy to be extended to multi-label segmentation, and for some less complicated scenarios, it can even get the segmented object through single line interaction. Experimental results and comparisons with other state-of-the-art methods demonstrate that our framework can efficiently and accurately extract foreground objects from background.
登录
查看更多内容
影响因子：
2.5
作者：
Hua Huang;Lei Zhang;Hong-Chao Zhang
通讯作者：
Hua Huang;Lei Zhang;Hong-Chao Zhang
影响因子：
10.6
作者：
Li, Chunming;Xu, Chenyang;Fox, Martin D.
通讯作者：
Fox, Martin D.
DOI：
10.1007/978-3-540-88690-7_20
发表时间：
2008-10
期刊：
--
影响因子：
--
作者：
Tae Hoon Kim;Kyoung Mu Lee;Sang Uk Lee
通讯作者：
Tae Hoon Kim;Kyoung Mu Lee;Sang Uk Lee
DOI：
10.1109/34.868688
发表时间：
2000-08-01
影响因子：
23.6
作者：
Shi, JB;Malik, J
通讯作者：
Malik, J
DOI：
10.1145/1576246.1531375
发表时间：
2009-07
期刊：
ACM SIGGRAPH 2009 papers
影响因子：
--
作者：
Jiangyu Liu;Jian Sun-;H. Shum
通讯作者：
Jiangyu Liu;Jian Sun-;H. Shum