US 12,437,566 B2
Picture annotation method, apparatus, electronic device, and storage medium
Yang Ding, Beijing (CN); and Ya Gao, Beijing (CN)
Assigned to BEIJING BAIDU NETCOM SCIENCE TECHNOLOGY CO., LTD., Beijing (CN)
Filed by BEIJING BAIDU NETCOM SCIENCE TECHNOLOGY CO., LTD., Beijing (CN)
Filed on Dec. 20, 2022, as Appl. No. 18/084,882.
Claims priority of application No. 202111575386.5 (CN), filed on Dec. 20, 2021.
Prior Publication US 2023/0119741 A1, Apr. 20, 2023
Int. Cl. G06V 20/70 (2022.01); G06F 40/169 (2020.01); G06T 7/70 (2017.01)
CPC G06V 20/70 (2022.01) [G06F 40/169 (2020.01); G06T 7/70 (2017.01); G06T 2207/30241 (2013.01)] 15 Claims
OG exemplary drawing
 
1. A picture annotation method, comprising:
starting an annotation tool to recognize a hierarchical relationship between candidate layers in the annotation tool;
acquiring an annotation event of a picture to be annotated;
distributing the annotation event to at least one target master candidate layer based on the hierarchical relationship;
distributing, for an arbitrary target master candidate layer, in the case where the arbitrary target master candidate layer is configured with a slave candidate layer, starting from the arbitrary target master candidate layer, the annotation event level by level to target slave candidate layers of at least one level of the arbitrary target master candidate layer;
processing the annotation event through a first target candidate layer, to which the annotation event is distributed, wherein the first target candidate layer includes the target master candidate layer and/or the target slave candidate layer; or
processing, for an arbitrary target master candidate layer, in the case where the arbitrary target master candidate layer is configured with no slave candidate layer, the annotation event through a second target candidate layer, to which the annotation event is distributed, wherein the second target candidate layer includes the target master candidate layer.