Visual Segmentation of Simple Objects for Robots
Visual Segmentation of Simple Objects for Robots
复制标题
DOI：
10.15607/rss.2011.vii.030
复制
发表时间：
2011-06
期刊：
影响因子：
--
通讯作者：
Ajay K. Mishra;Y. Aloimonos
中科院分区：
文献类型：
--
作者：
Ajay K. Mishra;Y. Aloimonos
自动分割任何大小的“简单”对象与其背景的能力对于主动代理（例如机器人）在真实的世界中有效地交互是重要的。最近，我们提出了一种算法[12]，以尺度不变的方式分割“简单”对象，给定对象内部的任何一个点。然而，在[12]中，没有提供选择“简单”对象内的点的策略。在本文中，我们提出了一个新的系统，自动选择不同的“简单”的对象在场景中的点，进行分割过程中所选择的点，并只输出对应于“简单”的对象在场景中的区域。针对分割问题提出的注意力机制首次利用了边界所有权的概念[17]。
The ability to automatically segment a “simple” object of any size from its background is important for an active agent (e.g. a robot) to interact effectively in the real world. Recently, we proposed an algorithm [12] to segment a “simple” object in a scale invariant manner, given a point anywhere inside that object. However, in [12], a strategy to select the point inside a “simple” object was not provided. In this paper, we propose a new system that automatically selects the points inside different “simple” objects in the scene, carries out the segmentation process for the selected points, and outputs only the regions corresponding to the “simple” objects in the scene. The proposed attention mechanism for the segmentation problem utilizes, for the first time, the concept of border ownership [17].