US 12,147,766 B2
Generating domain specific language expressions based on images
Shuhao Fu, San Jose, CA (US); Alexander Ngai, Irvine, CA (US); and Yueqi Li, San Jose, CA (US)
Assigned to Deere & Company, Moline, IL (US)
Filed by Deere & Company, Moline, IL (US)
Filed on Oct. 19, 2022, as Appl. No. 17/969,425.
Prior Publication US 2024/0134936 A1, Apr. 25, 2024
Prior Publication US 2024/0232532 A9, Jul. 11, 2024
Int. Cl. G06F 40/284 (2020.01); G06F 40/40 (2020.01); G06T 7/11 (2017.01); G06T 15/00 (2011.01); G06T 17/00 (2006.01)
CPC G06F 40/284 (2020.01) [G06F 40/40 (2020.01); G06T 7/11 (2017.01); G06T 15/00 (2013.01); G06T 17/00 (2013.01); G06T 2207/20021 (2013.01); G06T 2207/20081 (2013.01); G06T 2207/20084 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method for using a domain specific language (DSL) to generate training data, the method implemented using one or more processors and comprising:
processing data indicative of one or more ground truth images depicting a real plant using a trained image-to-DSL machine learning (ML) model to generate a first expression in the DSL that describes structure of the real plant, wherein the first expression includes a plurality of parameters;
processing the first expression to programmatically generate a plurality of synthetic DSL expressions, wherein each respective synthetic DSL expression describes structure of a respective synthetic plant for which one or more of the parameters has been altered from the first expression;
processing the plurality of synthetic DSL expressions using a renderer to create a plurality of three-dimensional (3D) synthetic plant models; and
generating one or more two-dimensional (2D) synthetic images that depict the plurality of 3D synthetic plant models in an area.