An algorithm for obstacle detection based on YOLO and light field camera

Rumin Zhang; YYifeng Yang; Wenyi Wang; Liaoyuan Zeng; Jianwen Chen; Sean McGrath

doi:10.1109/ICSensT.2018.8603600

An algorithm for obstacle detection based on YOLO and light field camera

Rumin Zhang, YYifeng Yang, Wenyi Wang, Liaoyuan Zeng, Jianwen Chen, Sean McGrath

Western Sydney University

Research output: Chapter in Book / Conference Paper › Conference Paper › peer-review

29 Citations (Scopus)

Abstract

This paper presents a novel obstacle detection algorithm in the indoor environment. The algorithm combines the YOLO object detection algorithm and the light field camera which is more simple than normal RGB-D sensor and acquires depth image and high-resolution images at the same in one exposure. The RGB Image rendered by the light filed camera is taken as an input of the YOLO model which was trained base on nearly 100 categories of common objects. According to the object information and the depth map, the obstacle was accurately calculated including its size and position. Experimental results demonstrate that the proposed method can provide higher detection accuracy under indoor environment.

Original language	English
Title of host publication	Proceedings of 2018 12th International Conference on Sensing Technology (ICST 2018), Limerick, Ireland, 4-6 December 2018
Publisher	IEEE
Pages	223-226
Number of pages	4
ISBN (Print)	9781538651476
DOIs	https://doi.org/10.1109/ICSensT.2018.8603600
Publication status	Published - 2018
Event	International Conference on Sensing Technology - Duration: 2 Dec 2019 → …

Publication series

Name
ISSN (Print)	2156-8065

Conference

Conference	International Conference on Sensing Technology
Period	2/12/19 → …

Access to Document

10.1109/ICSensT.2018.8603600

Cite this

@inproceedings{f8dc694207324f1798c04da599984137,

title = "An algorithm for obstacle detection based on YOLO and light field camera",

abstract = "This paper presents a novel obstacle detection algorithm in the indoor environment. The algorithm combines the YOLO object detection algorithm and the light field camera which is more simple than normal RGB-D sensor and acquires depth image and high-resolution images at the same in one exposure. The RGB Image rendered by the light filed camera is taken as an input of the YOLO model which was trained base on nearly 100 categories of common objects. According to the object information and the depth map, the obstacle was accurately calculated including its size and position. Experimental results demonstrate that the proposed method can provide higher detection accuracy under indoor environment.",

author = "Rumin Zhang and YYifeng Yang and Wenyi Wang and Liaoyuan Zeng and Jianwen Chen and Sean McGrath",

year = "2018",

doi = "10.1109/ICSensT.2018.8603600",

language = "English",

isbn = "9781538651476",

publisher = "IEEE",

pages = "223--226",

booktitle = "Proceedings of 2018 12th International Conference on Sensing Technology (ICST 2018), Limerick, Ireland, 4-6 December 2018",

note = "International Conference on Sensing Technology ; Conference date: 02-12-2019",

}

Zhang, R, Yang, YY, Wang, W, Zeng, L, Chen, J & McGrath, S 2018, An algorithm for obstacle detection based on YOLO and light field camera. in Proceedings of 2018 12th International Conference on Sensing Technology (ICST 2018), Limerick, Ireland, 4-6 December 2018. IEEE, pp. 223-226, International Conference on Sensing Technology, 2/12/19. https://doi.org/10.1109/ICSensT.2018.8603600

TY - GEN

T1 - An algorithm for obstacle detection based on YOLO and light field camera

AU - Zhang, Rumin

AU - Yang, YYifeng

AU - Wang, Wenyi

AU - Zeng, Liaoyuan

AU - Chen, Jianwen

AU - McGrath, Sean

PY - 2018

Y1 - 2018

N2 - This paper presents a novel obstacle detection algorithm in the indoor environment. The algorithm combines the YOLO object detection algorithm and the light field camera which is more simple than normal RGB-D sensor and acquires depth image and high-resolution images at the same in one exposure. The RGB Image rendered by the light filed camera is taken as an input of the YOLO model which was trained base on nearly 100 categories of common objects. According to the object information and the depth map, the obstacle was accurately calculated including its size and position. Experimental results demonstrate that the proposed method can provide higher detection accuracy under indoor environment.

AB - This paper presents a novel obstacle detection algorithm in the indoor environment. The algorithm combines the YOLO object detection algorithm and the light field camera which is more simple than normal RGB-D sensor and acquires depth image and high-resolution images at the same in one exposure. The RGB Image rendered by the light filed camera is taken as an input of the YOLO model which was trained base on nearly 100 categories of common objects. According to the object information and the depth map, the obstacle was accurately calculated including its size and position. Experimental results demonstrate that the proposed method can provide higher detection accuracy under indoor environment.

UR - https://hdl.handle.net/1959.7/uws:67113

U2 - 10.1109/ICSensT.2018.8603600

DO - 10.1109/ICSensT.2018.8603600

M3 - Conference Paper

SN - 9781538651476

SP - 223

EP - 226

BT - Proceedings of 2018 12th International Conference on Sensing Technology (ICST 2018), Limerick, Ireland, 4-6 December 2018

PB - IEEE

T2 - International Conference on Sensing Technology

Y2 - 2 December 2019

ER -

An algorithm for obstacle detection based on YOLO and light field camera

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this