US 12,272,140 B2
Method for detecting objects and their state, apparatus and storage medium applying method, and electronic device
Yi-Chen Wang, New Taipei (TW)
Assigned to HON HAI PRECISION INDUSTRY CO., LTD., New Taipei (TW)
Filed by HON HAI PRECISION INDUSTRY CO., LTD., New Taipei (TW)
Filed on Dec. 17, 2021, as Appl. No. 17/554,497.
Claims priority of application No. 202011566690.9 (CN), filed on Dec. 25, 2020.
Prior Publication US 2022/0207877 A1, Jun. 30, 2022
Int. Cl. G06V 20/50 (2022.01); G06T 7/70 (2017.01); G06V 10/50 (2022.01); G06V 10/774 (2022.01)
CPC G06V 20/50 (2022.01) [G06T 7/70 (2017.01); G06V 10/50 (2022.01); G06V 10/774 (2022.01); G06T 2207/20021 (2013.01); G06T 2207/20081 (2013.01); G06T 2207/30148 (2013.01)] 7 Claims
OG exemplary drawing
 
1. A method for detecting target objects in a container, the method comprising:
dividing an area where states of placement of the target objects in the container, to generate N number of sub-areas; wherein N being a positive integer;
obtaining an artificial intelligence model; wherein the artificial intelligence model is a model obtained by training according to a training images, and the training images are images of the sub-areas;
obtaining images taken by a camera corresponding to the sub-area; wherein each sub-area is provided with one camera;
inputting the images into the artificial intelligence model, and obtaining the states of placement of the target objects in the sub-areas;
wherein the dividing an area where states of placement of the target objects in the container, to generate N number of sub-areas comprises:
determining an area where the states of placement of the target objects in the container is recognized according to positions where the target objects are placed in the container; and
dividing the area to generate the N number of sub-areas; wherein the sub-areas comprise the area where the target objects are placed;
wherein the training images comprises an image of the target object is placed, an image where the target object is not placed, a pre-stored image when the target object is correctly placed, and a pre-stored image when the target object is incorrectly placed.