From contours to regions: An empirical evaluation
From contours to regions: An empirical evaluation
复制标题
DOI：
10.1109/cvpr.2009.5206707
复制
发表时间：
2009-06
期刊：
影响因子：
--
通讯作者：
Pablo Arbeláez;M. Maire;Charless C. Fowlkes;Jitendra Malik
中科院分区：
文献类型：
--
作者：
Pablo Arbeláez;M. Maire;Charless C. Fowlkes;Jitendra Malik
我们提出了一种通用的分组算法，从任何轮廓检测器的输出构建区域层次结构。我们的方法包括两个步骤，首先是面向分水岭变换（OWT），从等高线形成初始区域，然后是构建超度量等高线地图（UCM），定义分层分割。我们提供了广泛的实验评估来证明，当与高性能轮廓检测器相结合时，OWT-UCM算法产生了最先进的图像分割。这些分层分割可以通过用户指定的注释进一步细化。
We propose a generic grouping algorithm that constructs a hierarchy of regions from the output of any contour detector. Our method consists of two steps, an oriented watershed transform (OWT) to form initial regions from contours, followed by construction of an ultra-metric contour map (UCM) defining a hierarchical segmentation. We provide extensive experimental evaluation to demonstrate that, when coupled to a high-performance contour detector, the OWT-UCM algorithm produces state-of-the-art image segmentations. These hierarchical segmentations can optionally be further refined by user-specified annotations.