Overview
Currently, perceiving the planar objects (POP) in vision tasks is widely studied, e.g., planar object detection, planar object segmentation, planar object reconstruction, and planar object tracking. All of the corresponding tasks involve the same targets, i.e., plane. Due to its rich geometric information as a primitive element, POP has been explicitly explored for scene understanding and target analysis. With the help of analyzing planar objects, it is possible to develop applications in various downstream areas, including industry, art, security, education, commerce, and our daily life. Compared with common objects, POP focuses on class-agnostic targets, which have a nearly planar surface, even for a piece-wise shape.
Planar Object Segmentation
Planar Object Reconstruction
Planar Object Tracking
Citation
Zhicheng Zhang, Shengzhe Liu, Jufeng Yang. Multiple Planar Object Tracking. ICCV, 2023.
[PDF]
[HTML]
[Video]
[Project Page]
[Github]
[MPOT-3K Dataset]
[Demo]
Zhicheng Zhang, Song Chen, Zichuan Wang, Jufeng Yang. PlaneSeg: Building a Plug-in for Boosting Planar Region Segmentation. TNNLS, 2023.
[PDF]
[HTML]
[Project Page]
[Github]
[Demo]
Bibtex
@inproceedings{zhang2023mpot,
author={Zhang, Zhicheng and Liu, Shengzhe and Yang, Jufeng},
title={Multiple Planar Object Tracking},
booktitle={ICCV},
year={2023}
}
@article{zhang2023PlaneSeg,
author={Zhang, Zhicheng and Chen, Song and Wang, Zichuan and Yang, Jufeng},
title={PlaneSeg: Building a Plug-In for Boosting Planar Region Segmentation},
journal={IEEE Transactions on Neural Networks and Learning Systems},
year={2023}
}
