Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
APPARATUS AND METHOD FOR CLASSIFYING FAULTS OF RADAR SENSOR ON BASIS OF DEEP LEARNING BY USING CAMERA AND RADAR SENSOR
Document Type and Number:
WIPO Patent Application WO/2023/234567
Kind Code:
A1
Abstract:
The present invention relates to an apparatus for classifying faults of a radar sensor on the basis of deep learning by using a camera and a radar sensor, wherein the apparatus comprises: an input unit that receives, from a camera, an input of an image frame in which object boxes are displayed and receives radar information from a radar sensor; a correction unit that matches the radar information to an image in which the object boxes are displayed and displays radar points in the object boxes; a feature extraction unit that selects any one object box from among the object boxes present in the image, extracts features of an object, and preprocesses radar information of a radar point displayed in the object box; and an output unit that inputs the features of the object and the radar information into an artificial neural network model and outputs whether the radar point is normal. Accordingly, it may be classified whether a corresponding radar sensor is normal or faulty, by determining whether a radar point displayed in an image frame is normal.

Inventors:
HAN DONGSEOG (KR)
YOO MINWOO (KR)
SEO HYOJEONG (KR)
NING DIAN (KR)
Application Number:
PCT/KR2023/005742
Publication Date:
December 07, 2023
Filing Date:
April 27, 2023
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
KYUNGPOOK NAT UNIV IND ACADEMIC COOP FOUND (KR)
International Classes:
G01S7/40; G05B23/02; G06N3/08; G06V10/40
Foreign References:
KR20180068578A2018-06-22
KR20200010640A2020-01-31
US20070073473A12007-03-29
KR20210061597A2021-05-28
Other References:
JAE HEE KIM, HYUK-JIN YOON: "Research on Real-time Fault Diagnosis for Autonomous Public Transportation", JOURNAL OF THE KOREAN SOCIETY FOR RAILWAY, vol. 24, no. 2, 1 February 2021 (2021-02-01), pages 210 - 218, XP009551390, ISSN: 1738-6225, DOI: 10.7782/JKSR.2021.24.2.210
Attorney, Agent or Firm:
YUN, Kuisang (KR)
Download PDF: