Exploring Surround-View Fisheye Camera 3D Object Detection

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Li, Changcai, Lin, Wenwei, Hou, Zuoxun, Chen, Gang, Zhang, Wei, Zhou, Huihui, Zheng, Weishi
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915634808356864
author Li, Changcai
Lin, Wenwei
Hou, Zuoxun
Chen, Gang
Zhang, Wei
Zhou, Huihui
Zheng, Weishi
author_facet Li, Changcai
Lin, Wenwei
Hou, Zuoxun
Chen, Gang
Zhang, Wei
Zhou, Huihui
Zheng, Weishi
contents In this work, we explore the technical feasibility of implementing end-to-end 3D object detection (3DOD) with surround-view fisheye camera system. Specifically, we first investigate the performance drop incurred when transferring classic pinhole-based 3D object detectors to fisheye imagery. To mitigate this, we then develop two methods that incorporate the unique geometry of fisheye images into mainstream detection frameworks: one based on the bird's-eye-view (BEV) paradigm, named FisheyeBEVDet, and the other on the query-based paradigm, named FisheyePETR. Both methods adopt spherical spatial representations to effectively capture fisheye geometry. In light of the lack of dedicated evaluation benchmarks, we release Fisheye3DOD, a new open dataset synthesized using CARLA and featuring both standard pinhole and fisheye camera arrays. Experiments on Fisheye3DOD show that our fisheye-compatible modeling improves accuracy by up to 6.2% over baseline methods.
format Preprint
id arxiv_https___arxiv_org_abs_2511_18695
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Exploring Surround-View Fisheye Camera 3D Object Detection
Li, Changcai
Lin, Wenwei
Hou, Zuoxun
Chen, Gang
Zhang, Wei
Zhou, Huihui
Zheng, Weishi
Computer Vision and Pattern Recognition
I.2.10; I.4.8
In this work, we explore the technical feasibility of implementing end-to-end 3D object detection (3DOD) with surround-view fisheye camera system. Specifically, we first investigate the performance drop incurred when transferring classic pinhole-based 3D object detectors to fisheye imagery. To mitigate this, we then develop two methods that incorporate the unique geometry of fisheye images into mainstream detection frameworks: one based on the bird's-eye-view (BEV) paradigm, named FisheyeBEVDet, and the other on the query-based paradigm, named FisheyePETR. Both methods adopt spherical spatial representations to effectively capture fisheye geometry. In light of the lack of dedicated evaluation benchmarks, we release Fisheye3DOD, a new open dataset synthesized using CARLA and featuring both standard pinhole and fisheye camera arrays. Experiments on Fisheye3DOD show that our fisheye-compatible modeling improves accuracy by up to 6.2% over baseline methods.
title Exploring Surround-View Fisheye Camera 3D Object Detection
topic Computer Vision and Pattern Recognition
I.2.10; I.4.8
url https://arxiv.org/abs/2511.18695