FatRegion: A Fast, Adaptive, Tree-structured Region Extraction Approach

CORC > 自动化研究所 > 中国科学院自动化研究所 > 模式识别国家重点实验室 > 视频内容安全团队

	FatRegion: A Fast, Adaptive, Tree-structured Region Extraction Approach
	Junliang Xing; Weiming Hu; Haozhou Ai; Shuicheng Yan
刊名	IEEE Transactions on Circuits and Systems for Video Technology
	2016
卷号	PP 期号:99 页码:1-17
关键词	Object Classification Region Extraction Superpixel Object Segmentation Object Tracking
英文摘要	Coherent image regions can be used as good features for many computer vision tasks, such as object tracking, segmentation, and recognition. Most of previous region extraction methods, however, are not suitable for online applications because of either their heavy computations or unsatisfactory results. We propose a seed based region growing and merging approach to generate simultaneously coherent and discriminative image regions. We present a quad-tree based seed initialization algorithm to adaptively place seeds into different image areas and then grow them into regions by a color- and edge-guided growing procedure. To merge these regions in different levels, we propose to use the generalized boundary strength to measure the quality of region merging result. And we present a region merging algorithm of linear time complexity to perform efficient and effective region merging. Overall, our new approach simultaneously holds these advantages: 1) it is extremely fast with linear complexity both in time and space, which takes less than 50 milliseconds to process a HVGA image; 2) it can give a direct control of the region number and well adapt to image regions with various sizes and shapes; 3) it provides a tree-structured representation of the regions, thus can model the image from multiple scales. We evaluate the proposed approach on the standard benchmarks with extensive comparisons to the state-of-the-art methods. Experimental results demonstrate its good comprehensive performances. Example applications using the extracted regions as features for online object tracking and multi-class object segmentation also exhibit its potential for many computer vision tasks.
URL标识	查看原文
会议地点	IEEE
会议日期	2016
内容类型	期刊论文
源URL	[http://ir.ia.ac.cn/handle/173211/13723]
专题	自动化研究所_模式识别国家重点实验室_视频内容安全团队
作者单位	National Laboratory of Pattern Recognition, Institute of Automation, Chinese Academy of Sciences
推荐引用方式 GB/T 7714	Junliang Xing,Weiming Hu,Haozhou Ai,et al. FatRegion: A Fast, Adaptive, Tree-structured Region Extraction Approach[J]. IEEE Transactions on Circuits and Systems for Video Technology,2016,PP(99):1-17.
APA	Junliang Xing,Weiming Hu,Haozhou Ai,&Shuicheng Yan.(2016).FatRegion: A Fast, Adaptive, Tree-structured Region Extraction Approach.IEEE Transactions on Circuits and Systems for Video Technology,PP(99),1-17.
MLA	Junliang Xing,et al."FatRegion: A Fast, Adaptive, Tree-structured Region Extraction Approach".IEEE Transactions on Circuits and Systems for Video Technology PP.99(2016):1-17.