Automotive radar image segmentation with frame fusion

Yang Xiao; Liam Daniel; Scott Cassidy; Sukhjit Pooni; Edward Hoare; Anum Pirkani; Mike Cherniakov; Marina Gashinova

doi:10.1049/icp.2022.2323

Automotive radar image segmentation with frame fusion

Yang Xiao, Liam Daniel, Scott Cassidy, Sukhjit Pooni, Edward Hoare, Anum Pirkani, Mike Cherniakov, Marina Gashinova

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

Image segmentation on automotive radar imagery is the key technique for identifying the passable and impassable regions for path planning in autonomous or assistive driving. The availability of consecutive frames which measure the driving scene shifted along with the timeline enables improved segmentation on radar imagery. The frame fusion on automotive radar map is implemented as a two-step procedure: 1) The pixel-to-pixel mapping between consecutive frames is achieved based on an inertial measurement unit (IMU); 2) The information fusion of consecutive frames is achieved based on the Kalman filter. The frame fusion operation leads to correct classification of the initially “unknown” regions and overall improves the confidence of classification compared to single frame segmentation. The segmentation results with frame fusion are presented and compared with the results of single frame segmentation.

Original language	English
Title of host publication	International Conference on Radar Systems (RADAR 2022)
Publisher	Institution of Engineering and Technology (IET)
Number of pages	5
ISBN (Electronic)	9781839537776
DOIs	https://doi.org/10.1049/icp.2022.2323
Publication status	Published - 7 Feb 2022
Event	International Conference on Radar Systems (RADAR 2022) - Murrayfield Stadium, Edinburgh, United Kingdom Duration: 24 Oct 2022 → 27 Oct 2022 Conference number: CP804

Publication series

Name	International Conference on Radar Systems

Conference

Conference	International Conference on Radar Systems (RADAR 2022)
Abbreviated title	RADAR 2022
Country/Territory	United Kingdom
City	Edinburgh
Period	24/10/22 → 27/10/22

Bibliographical note

Acknowledgments:
This work is supported by Innovate UK grant 104268 and is part of the project "CORTEX-Cognitive Real-Time System for Autonomous Vehicles".

Access to Document

10.1049/icp.2022.2323

CORTEX
Gashinova, M. (Principal Investigator), Cherniakov, M. (Co-Investigator) & Windridge, D. (Co-Investigator)
TECHNOLOGY STRATEGY BOARD
1/03/19 → 31/03/22
Project: Other Government Departments

Cite this

Xiao, Y., Daniel, L., Cassidy, S., Pooni, S., Hoare, E., Pirkani, A., Cherniakov, M., & Gashinova, M. (2022). Automotive radar image segmentation with frame fusion. In International Conference on Radar Systems (RADAR 2022) Article 10038922 (International Conference on Radar Systems). Institution of Engineering and Technology (IET). https://doi.org/10.1049/icp.2022.2323

@inproceedings{0ef5263beb144f2d88f8acd9711b561f,

title = "Automotive radar image segmentation with frame fusion",

abstract = "Image segmentation on automotive radar imagery is the key technique for identifying the passable and impassable regions for path planning in autonomous or assistive driving. The availability of consecutive frames which measure the driving scene shifted along with the timeline enables improved segmentation on radar imagery. The frame fusion on automotive radar map is implemented as a two-step procedure: 1) The pixel-to-pixel mapping between consecutive frames is achieved based on an inertial measurement unit (IMU); 2) The information fusion of consecutive frames is achieved based on the Kalman filter. The frame fusion operation leads to correct classification of the initially “unknown” regions and overall improves the confidence of classification compared to single frame segmentation. The segmentation results with frame fusion are presented and compared with the results of single frame segmentation.",

author = "Yang Xiao and Liam Daniel and Scott Cassidy and Sukhjit Pooni and Edward Hoare and Anum Pirkani and Mike Cherniakov and Marina Gashinova",

note = "Acknowledgments: This work is supported by Innovate UK grant 104268 and is part of the project {"}CORTEX-Cognitive Real-Time System for Autonomous Vehicles{"}.; International Conference on Radar Systems (RADAR 2022), RADAR 2022 ; Conference date: 24-10-2022 Through 27-10-2022",

year = "2022",

month = feb,

day = "7",

doi = "10.1049/icp.2022.2323",

language = "English",

series = "International Conference on Radar Systems",

publisher = "Institution of Engineering and Technology (IET)",

booktitle = "International Conference on Radar Systems (RADAR 2022)",

}

Xiao, Y, Daniel, L, Cassidy, S, Pooni, S, Hoare, E , Pirkani, A , Cherniakov, M & Gashinova, M 2022, Automotive radar image segmentation with frame fusion. in International Conference on Radar Systems (RADAR 2022)., 10038922, International Conference on Radar Systems, Institution of Engineering and Technology (IET), International Conference on Radar Systems (RADAR 2022), Edinburgh, United Kingdom, 24/10/22. https://doi.org/10.1049/icp.2022.2323

TY - GEN

T1 - Automotive radar image segmentation with frame fusion

AU - Xiao, Yang

AU - Daniel, Liam

AU - Cassidy, Scott

AU - Pooni, Sukhjit

AU - Hoare, Edward

AU - Pirkani, Anum

AU - Cherniakov, Mike

AU - Gashinova, Marina

N1 - Conference code: CP804

PY - 2022/2/7

Y1 - 2022/2/7

N2 - Image segmentation on automotive radar imagery is the key technique for identifying the passable and impassable regions for path planning in autonomous or assistive driving. The availability of consecutive frames which measure the driving scene shifted along with the timeline enables improved segmentation on radar imagery. The frame fusion on automotive radar map is implemented as a two-step procedure: 1) The pixel-to-pixel mapping between consecutive frames is achieved based on an inertial measurement unit (IMU); 2) The information fusion of consecutive frames is achieved based on the Kalman filter. The frame fusion operation leads to correct classification of the initially “unknown” regions and overall improves the confidence of classification compared to single frame segmentation. The segmentation results with frame fusion are presented and compared with the results of single frame segmentation.

AB - Image segmentation on automotive radar imagery is the key technique for identifying the passable and impassable regions for path planning in autonomous or assistive driving. The availability of consecutive frames which measure the driving scene shifted along with the timeline enables improved segmentation on radar imagery. The frame fusion on automotive radar map is implemented as a two-step procedure: 1) The pixel-to-pixel mapping between consecutive frames is achieved based on an inertial measurement unit (IMU); 2) The information fusion of consecutive frames is achieved based on the Kalman filter. The frame fusion operation leads to correct classification of the initially “unknown” regions and overall improves the confidence of classification compared to single frame segmentation. The segmentation results with frame fusion are presented and compared with the results of single frame segmentation.

U2 - 10.1049/icp.2022.2323

DO - 10.1049/icp.2022.2323

M3 - Conference contribution

T3 - International Conference on Radar Systems

BT - International Conference on Radar Systems (RADAR 2022)

PB - Institution of Engineering and Technology (IET)

T2 - International Conference on Radar Systems (RADAR 2022)

Y2 - 24 October 2022 through 27 October 2022

ER -

Automotive radar image segmentation with frame fusion

Abstract

Publication series

Conference

Bibliographical note

Access to Document

Fingerprint

Projects

CORTEX

Cite this