US 12,260,540 B2
Material completeness detection method and apparatus, and storage medium
Hao Li, Zhengzhou (CN); Xinyu Yan, Zhengzhou (CN); Gen Liu, Zhengzhou (CN); Haoqi Wang, Zhengzhou (CN); Zhongshang Zhai, Zhengzhou (CN); Bing Li, Zhengzhou (CN); Yuyan Zhang, Zhengzhou (CN); and Yan Liu, Zhengzhou (CN)
Assigned to ZHENGZHOU UNIVERSITY OF LIGHT INDUSTRY, Zhengzhou (CN)
Filed by ZHENGZHOU UNIVERSITY OF LIGHT INDUSTRY, Zhengzhou (CN)
Filed on May 20, 2022, as Appl. No. 17/749,195.
Prior Publication US 2023/0377123 A1, Nov. 23, 2023
Int. Cl. G06T 7/00 (2017.01); G05B 19/418 (2006.01); G06V 10/82 (2022.01)
CPC G06T 7/0008 (2013.01) [G05B 19/41875 (2013.01); G06V 10/82 (2022.01); G05B 2219/32335 (2013.01); G06T 2200/24 (2013.01); G06T 2207/20084 (2013.01)] 7 Claims
OG exemplary drawing
 
1. A material completeness detection method, for detecting whether materials of a target object in a physical production line are complete, comprising:
inputting an image of the target object into a material completeness detection algorithm to acquire a first detection result;
inputting a virtual model of the target object in a virtual production line into the material completeness detection algorithm to acquire a second detection result, wherein the virtual production line is a digital twin (DT) of the physical production line; and
acquiring a material completeness detection result of the target object based on the first detection result and the second detection result;
the material completeness detection algorithm comprises:
a backbone feature extraction network, configured to receive the image, and extract a backbone feature of the image to generate a 160*160 feature layer, an 80*80 feature layer, a 40*40 feature layer, and a 20*20 feature layer;
an enhanced feature extraction network, configured to extract an enhanced feature from the input 160*160 feature layer, 80*80 feature layer, 40*40 feature layer and 20*20 feature layer to generate a 160*160*128 enhanced feature layer, an 80*80*256 enhanced feature layer, a 40*40*512 enhanced feature layer, and a 20*20*1,024 enhanced feature layer; and
an output network, configured to output a detection result based on the 160*160*128 enhanced feature layer, the 80*80*256 enhanced feature layer, the 40*40*512 enhanced feature layer, and the 20*20*1,024 enhanced feature layer;
the material completeness detection algorithm is defined as any logic for determining if production of a material is complete.