US 12,243,211 B2
Method to train a neural network to detect a tool status from images, method of machining and/or manufacturing, and installation
Benjamin Samuel Lutz, Munich (DE); and Daniel Regulin, Munich (DE)
Assigned to Siemens Aktiengesellschaft, Munich (DE)
Filed by Benjamin Samuel Lutz, Munich (DE); and Daniel Regulin, Munich (DE)
Filed on Aug. 27, 2021, as Appl. No. 17/460,155.
Claims priority of application No. 20193430 (EP), filed on Aug. 28, 2020.
Prior Publication US 2022/0067913 A1, Mar. 3, 2022
Int. Cl. G06T 7/00 (2017.01); G05B 19/4065 (2006.01); G06N 3/08 (2023.01)
CPC G06T 7/0004 (2013.01) [G05B 19/4065 (2013.01); G06N 3/08 (2013.01); G06T 2207/10056 (2013.01); G06T 2207/20081 (2013.01)] 12 Claims
OG exemplary drawing
 
1. A method for training a neural network to recognize a machining tool condition based on image data, the method comprising:
training the neural network to recognize the machining tool condition of a first tool type; and
applying image data of a second tool type, the applied image data being subjected to image processing, via which the image data of the second tool type is converted into image data of the first tool type,
wherein the neural network is trained based on the converted image data,
wherein the image data of the second tool type is converted such that a shape of the second tool type is converted into a shape of the first tool type, and
wherein the image data of the second tool type is assigned a respective machining tool condition that is applied in each case jointly with the image data of the second tool type to train the neural network.