	US 12,094,310 B2
	Method and system for detecting fire and smoke
Michael Michael, Singapore (SG); and Chuan Hun Ye, Singapore (SG)
Assigned to HENDRICKS CORP PTE LTD, Singapore (SG)
Appl. No. 17/604,190
Filed by HENDRICKS CORP PTE LTD, Singapore (SG)
PCT Filed Apr. 17, 2019, PCT No. PCT/SG2019/050216 § 371(c)(1), (2) Date Oct. 15, 2021, PCT Pub. No. WO2020/214084, PCT Pub. Date Oct. 22, 2020.
Prior Publication US 2022/0198801 A1, Jun. 23, 2022
Int. Cl. G06K 9/00 (2022.01); G06T 7/12 (2017.01); G06T 7/269 (2017.01); G06V 10/44 (2022.01); G06V 20/00 (2022.01); G06V 20/50 (2022.01); G08B 17/10 (2006.01)

CPC G08B 17/10 (2013.01) [G06T 7/12 (2017.01); G06T 7/269 (2017.01); G06V 10/44 (2022.01); G06V 20/00 (2022.01); G06V 20/50 (2022.01); G06T 2207/10016 (2013.01); G06T 2207/10024 (2013.01)]

14 Claims

1. A smoke detection system comprising:

a processor, a memory and instructions stored on the memory and executable by the processor to:

receive a sequence of images from a plurality of cameras;

sampling the sequence of images at a certain interval;

process each of the sampled images to form a first processed image;

extract Binary Large Objects (BLOBs) from first processed image;

analyse the BLOB for smoke;

determine smoke in response to analysing smoke from the BLOBs;

process each of the sampled images to form a second processed image;

extract contours of objects from the second processed image;

analyse the contours of objects for fire; and

determine fire in response to analysing fire from the contours of objects,

wherein the instruction to analyse the contours of objects for fire comprises instructions to:

determine an average of X and Y coordinates for all contour pixels;

determine a distance, D_n, between each contour pixel with reference to the average of X and Y coordinates for all contour pixels;

determine an average distance, D_A, of all contour pixels with reference to the average of X and Y coordinates for all contour pixels;

determine a roundness index, T, is determined by applying the following expression,

T=1−Σ^{N/n=1 1DNA−Dn1/N}

where N is the total amount of pixels in the contour, and

remove contours with T<0.9.