跳到主要导航 跳到搜索 跳到主要内容

Window adaptive cost aggregation method for stereo correspondence

  • Jing Luo
  • , Zuren Feng
  • , Na Lu
  • Xi'an Jiaotong University

科研成果: 书/报告/会议事项章节会议稿件同行评审

摘要

Stereo matching is an active research area in computer vision field, and a lot of excellent researchers have focused on this challenging topic. For stereo matching, Census is proven to be a simple matching algorithm with high accuracy. But how to choose the aggregation window remains a critical problem. This paper presents a new approach to choose the size of aggregation window based on texture and the disparity boundary near the pixel to be matched, which does not increase much extra computational complexity. Experiment results show that the proposed method is effective to improve the matching result.

源语言英语
主期刊名Electronics, Communications and Networks IV - Proceedings of the 4th International Conference on Electronics, Communications and Networks, CECNet2014
编辑Amir Hussain, Mirjana Ivanovic
出版商CRC Press/Balkema
299-303
页数5
ISBN(印刷版)9781138028302
DOI
出版状态已出版 - 2015

出版系列

姓名Electronics, Communications and Networks IV - Proceedings of the 4th International Conference on Electronics, Communications and Networks, CECNet2014
1

学术指纹

探究 'Window adaptive cost aggregation method for stereo correspondence' 的科研主题。它们共同构成独一无二的指纹。

引用此