US 12,460,389 B2
Shovel and construction system
Takashi Nishi, Chiba (JP)
Assigned to SUMITOMO CONSTRUCTION MACHINERY CO., LTD, Tokyo (JP)
Filed by SUMITOMO CONSTRUCTION MACHINERY CO., LTD., Tokyo (JP)
Filed on Sep. 29, 2021, as Appl. No. 17/449,317.
Application 17/449,317 is a continuation of application No. PCT/JP2020/014696, filed on Mar. 30, 2020.
Claims priority of application No. 2019-069472 (JP), filed on Mar. 30, 2019.
Prior Publication US 2022/0018096 A1, Jan. 20, 2022
Int. Cl. E02F 9/26 (2006.01); E02F 9/22 (2006.01); G06F 18/24 (2023.01); G06F 18/2431 (2023.01); G06N 3/04 (2023.01); G06N 3/045 (2023.01)
CPC E02F 9/261 (2013.01) [E02F 9/226 (2013.01); G06F 18/2431 (2023.01); G06N 3/045 (2023.01)] 8 Claims
OG exemplary drawing
 
1. A shovel comprising:
a lower traveling structure;
an upper swing structure rotatably mounted on the lower traveling structure; and
an information obtaining device that obtains information related to construction, the information obtaining device including a sensor that captures an image related to a construction site; and
processing circuitry that inputs image data of the captured image to a learned model subjected to machine learning to:
process the input image data to identify types of objects included in the captured image,
predict probabilities of presence of the objects of the identified types,
predict probabilities of assumed scenes of the construction site based on the predicted probabilities of the objects of the identified types and a relative positional relationship between the objects of the identified types, and
determine a scene of the construction site based on the predicted probabilities of the assumed scenes,
wherein the processing circuitry further calculates a danger level of the determined scene of the construction site based on a type of the determined scene and generates an image to be displayed by superimposing the predicted probabilities of presence of the objects of the identified types, a predicted probability of the type of the determined scene, and the calculated danger level on the captured image.