US 12,394,173 B2
Facilitating annotation of objects
Thalanayar Muthukumar, Plano, TX (US); Volodya Grancharov, Solna (SE); and Marc Couch, Frisco, TX (US)
Assigned to TELEFONAKTIEBOLAGET LM ERICSSON (PUBL), Stockholm (SE)
Appl. No. 17/917,332
Filed by Telefonaktiebolaget LM Ericsson (publ), Stockholm (SE)
PCT Filed Apr. 6, 2020, PCT No. PCT/EP2020/059772
§ 371(c)(1), (2) Date Oct. 6, 2022,
PCT Pub. No. WO2021/204350, PCT Pub. Date Oct. 14, 2021.
Prior Publication US 2023/0154136 A1, May 18, 2023
Int. Cl. G06V 10/20 (2022.01); G06V 10/26 (2022.01); G06V 10/44 (2022.01); G06V 10/56 (2022.01); G06V 20/70 (2022.01)
CPC G06V 10/255 (2022.01) [G06V 10/273 (2022.01); G06V 10/44 (2022.01); G06V 10/56 (2022.01); G06V 20/70 (2022.01)] 8 Claims
OG exemplary drawing
 
1. A method of facilitating annotation of an object in a sequence of images, the method comprising:
marking the object in a first image in the sequence of images;
performing an edge segmentation, wherein performing the edge segmentation comprises producing a first binary image by applying an edge detector to the first image or to a first version of the first image;
using the first binary image to create a first edge mask;
performing a color segmentation, wherein performing the color segmentation comprises:
calculating a first color value using a first set of pixels from the first image or from a second version of the first image, wherein the first set of pixels correspond to the object;
calculating a second color value using a second set of pixels from the first image or from the second version of the first image, wherein the second set of pixels correspond to visual data in the first image other than the object;
creating a first color mask using the second color value; and
determining that a difference between the first color value and the second color value satisfies a threshold condition;
suppressing visual data other than the object, wherein suppressing visual data other than the object comprises masking the first image using both the first edge mask and the first color mask as a result of determining that the difference satisfies the threshold condition; and
aligning a marking with the object in a second image of the sequence, wherein the second image follows the first image in the sequence, wherein
the color segmentation is performed in parallel with the edge segmentation, and
the marked object in the second image is more distinguishable from its surroundings after the color segmentation and the edge segmentation is performed.