US 12,489,964 B2
Path planning
Li Cao, Beijing (CN); Ze Ming Zhao, Beijing (CN); Zhan Wei Wang, Xi'an (CN); and Xiao Tian Xu, Chang De (CN)
Assigned to International Business Machines Corporation, Armonk, NY (US)
Filed by International Business Machines Corporation, Armonk, NY (US)
Filed on Jul. 27, 2021, as Appl. No. 17/386,003.
Prior Publication US 2023/0036851 A1, Feb. 2, 2023
Int. Cl. G06F 18/213 (2023.01); G06N 20/00 (2019.01); G06T 7/20 (2017.01); G06V 20/00 (2022.01); H04N 23/61 (2023.01)
CPC H04N 23/61 (2023.01) [G06N 20/00 (2019.01); G06T 7/20 (2013.01); G06V 20/00 (2022.01)] 15 Claims
OG exemplary drawing
 
1. A computer-implemented method, comprising:
determining, by one or more processors, a kernel size approximating a size of a target image from an image capturing device;
selecting, by one or more processors, a target discriminator from a set of discriminators with different kernel sizes selected as a set for capturing objects at different sizes, wherein the target discriminator is selected based on the kernel size;
determining, by the one or more processors, using the target discriminator, a confidence of the target image, the confidence indicating whether the target image contains a target object to be captured;
responsive to the confidence of the target image being lower than a threshold confidence, determining, by the one or more processors, a movement indication is needed to indicate a path for moving the image capturing device;
determining a plurality of candidate movement indications in directions from a current point along which the image capturing device could move; and
determining, by the one or more processors, a plurality of weights associated with the plurality of candidate movement indications, a weight of the plurality of weights being associated with a possibility that the image capturing device captures an image of the target object after moving according to a corresponding candidate movement indication.