US 12,136,224 B2
Segmentation method
Niclas Danielsson, Lund (SE); Xing Danielsson Fan, Lund (SE); and Axel Keskikangas, Lund (SE)
Assigned to AXIS AB, Lund (SE)
Filed by Axis AB, Lund (SE)
Filed on Jul. 14, 2022, as Appl. No. 17/864,400.
Claims priority of application No. 21198349 (EP), filed on Sep. 22, 2021.
Prior Publication US 2023/0086993 A1, Mar. 23, 2023
Int. Cl. G06K 9/34 (2006.01); G06T 7/11 (2017.01); G06T 7/136 (2017.01); G06V 10/82 (2022.01); H04N 7/18 (2006.01)
CPC G06T 7/136 (2017.01) [G06T 7/11 (2017.01); G06V 10/82 (2022.01); H04N 7/18 (2013.01)] 14 Claims
OG exemplary drawing
 
1. A method of generating a segmentation outcome which indicates individual instances of one or more object classes for an image in a sequence of images, the method comprising:
a. determining a coherent region of the image,
b. processing the image to determine a tensor representing pixel-specific confidence scores for one or more object classes,
c. generating a series of temporary segmentation masks for the coherent region, wherein each temporary segmentation mask is generated by interpreting the tensor with respect to a single object class using a different temporary confidence score threshold,
d. evaluating the series of temporary segmentation masks to determine if an object mask condition is met by one or more of the temporary segmentation masks,
e. in case the object mask condition is met by one or more of the temporary segmentation masks, setting the temporary confidence score threshold used for generating one of the one or more temporary segmentation masks as a final confidence score threshold for the pixels of the temporary segmentation mask,
f. in case the object mask condition is not met, setting a default confidence score threshold as a final confidence score threshold for the coherent region,
g. generating a final segmentation outcome for the image, wherein a part of the final segmentation outcome covering the coherent region is generated by interpreting the tensor using the final confidence score threshold.