Robust people indoor localization with omnidirectional cameras using a Grid of Spatial-Aware Classifiers

作者：

Highlights：

• Only point-based annotation needed: faster and cheaper creation of databases.

• Robust to strong geometric distortions introduced by omnidirectional cameras.

• The system can work in real-time without specialized hardware, such as GPUs.

• New omnidirectional database for people detection (PIROPO) publicly available.

摘要

•Only point-based annotation needed: faster and cheaper creation of databases.•Robust to strong geometric distortions introduced by omnidirectional cameras.•The system can work in real-time without specialized hardware, such as GPUs.•New omnidirectional database for people detection (PIROPO) publicly available.

论文关键词：Spatial grid,Spatial-aware classifiers,Omnidirectional cameras,Fish-eye,Indoor localization,People detection,Object detection,Machine learning

论文评审过程：Received 24 March 2020, Revised 13 November 2020, Accepted 3 January 2021, Available online 12 January 2021, Version of Record 27 January 2021.

论文官网地址：https://doi.org/10.1016/j.image.2021.116135