US 11,951,562 B2
Laser processing system
Yohei Kobayashi, Tokyo (JP); Shuntaro Tani, Tokyo (JP); and Yutsuki Aoyagi, Tokyo (JP)
Assigned to THE UNIVERSITY OF TOKYO, Tokyo (JP)
Filed by THE UNIVERSITY OF TOKYO, Tokyo (JP)
Filed on Feb. 22, 2019, as Appl. No. 16/283,130.
Claims priority of application No. 2018-030874 (JP), filed on Feb. 23, 2018.
Prior Publication US 2019/0262936 A1, Aug. 29, 2019
Int. Cl. B23K 26/03 (2006.01); B23K 26/0622 (2014.01); B23K 26/064 (2014.01); B23K 26/36 (2014.01); B23K 26/362 (2014.01); B23K 26/70 (2014.01); G06N 3/08 (2023.01)
CPC B23K 26/032 (2013.01) [B23K 26/0624 (2015.10); B23K 26/064 (2015.10); B23K 26/36 (2013.01); B23K 26/362 (2013.01); B23K 26/705 (2015.10); G06N 3/08 (2013.01)] 7 Claims
OG exemplary drawing
 
1. A laser processing system equipped with a processing laser beam irradiation device configured to irradiate a processing object with a processing laser beam and perform ablation processing, the laser processing system comprising:
a camera configured to obtain an ablation image of a processed portion of the processing object based on scattered light from the processed portion during processing of the processing object with the processing laser beam; and
a processor programmed to estimate an ablation volume by inputting the ablation image into a deep learning model that specifies a relationship between the ablation image and the ablation volume and outputs the estimated ablation volume,
wherein the deep learning model has been trained by outputting a processing parameter based on the input ablation image, and applying a loss function to the deep learning model based on actual values of the processing parameter, and the deep learning model estimates the ablation volume based on the processing parameter.