Multi-object segmentation by stereo mismatch

被引：5

作者：

Sun, Wei ^{[1
]}

Spackman, Stephen P. ^{[2
]}

机构：

[1] Intel Corp, Santa Clara, CA 95054 USA

[2] Quantum Corp, San Jose, CA 95110 USA

来源：

MACHINE VISION AND APPLICATIONS | 2009年 / 20卷 / 06期

关键词：

Object segmentation; Scene decomposition; Background models; Stereo mismatch; Disparity contours;

D O I：

10.1007/s00138-008-0127-1

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

A new stereo mismatch based foreground object segmentation method is described. It efficiently locates objects over a wide range of depths against backgrounds of known 3D geometry, even in the presence of rapidly changing lighting and dynamic textures, such as projected video. Not relying on full stereo reconstruction, it is fast enough in software for some real-time applications, robust to camera quality, and requires little parameter tuning. Experimental results validate the approach, demonstrating its ability to simultaneously distinguish multiple objects in a complex scene, even when close together or partially occluded.

引用

页码：339 / 352

页数：14