US 12,303,729 B2
Fire detection system
Jongsong Park, Seongnam-si (KR); Ja-Yeon Jeong, Seongnam-si (KR); Young Kyoo Hwang, Suwon-si (KR); Jiheon Park, Suwon-si (KR); Hotaek Han, Siheung-si (KR); Jai Hoon Park, Yongin-si (KR); and Jung-Bae Kim, Seoul (KR)
Assigned to ALCHERAX, INC., Sunnyvale, CA (US)
Appl. No. 17/768,554
Filed by AlcheraX, Inc., Sunnyvale, CA (US)
PCT Filed Jan. 29, 2020, PCT No. PCT/KR2020/001357
§ 371(c)(1), (2) Date Apr. 13, 2022,
PCT Pub. No. WO2021/091021, PCT Pub. Date May 14, 2021.
Claims priority of application No. 10-2019-0142797 (KR), filed on Nov. 8, 2019.
Prior Publication US 2023/0201642 A1, Jun. 29, 2023
Int. Cl. A62C 3/02 (2006.01); G06T 5/70 (2024.01); G06T 5/73 (2024.01); G06T 7/215 (2017.01); G06T 7/246 (2017.01); G06V 10/62 (2022.01); G06V 10/764 (2022.01); G06V 10/774 (2022.01); G06V 20/10 (2022.01)
CPC A62C 3/0271 (2013.01) [G06T 5/70 (2024.01); G06T 5/73 (2024.01); G06T 7/215 (2017.01); G06T 7/246 (2017.01); G06V 10/62 (2022.01); G06V 10/764 (2022.01); G06V 10/774 (2022.01); G06V 20/188 (2022.01)] 7 Claims
OG exemplary drawing
 
1. A fire detecting system comprising:
an optical flow detecting module for estimating a motion of a fire element from an input image, and determining a first candidate area in the input image;
an image feature detecting module for recognizing an image feature of the fire element from the input image, and determining a second candidate area in the input image;
a candidate area analyzing module for determining a suspicion area in which generation of fire is suspected in the input image based on the first candidate area and the second candidate area;
an ignition point estimating module for estimating a position of an ignition point in the suspicion area;
a fire classifying module for calculating classifying scores obtained by predicting whether a fire is generated in the suspicion area;
a temporal analysis module for determining whether a fire is generated based on the position of the ignition point and the classifying scores; and
a stabilizing module for providing a shaking corrected image obtained by correcting shaking of the input image to the optical flow detecting module and the image feature detecting module,
wherein
the candidate area analyzing module outputs suspicion area information including at least one of an optical flow vector and an optical flow size showing an optical flow approximating a position of the suspicion area, a size of the suspicion area, a motion of the fire element in the input image, a soft mask indicating flame predicted from the input image, and a soft mask indicating predicted smoke in the input image, and
wherein
the stabilizing module includes
a motion vector extracting module for extracting a motion vector from the input image, and
a temporal smoothing module for performing a temporal smoothing process on the input image by using the motion vector.