US 12,266,090 B2
Systems and methods for detecting defects in powder manufactured components
Dominic N. Nwoke, Spanaway, WA (US); Robert W. Grube, Edmonds, WA (US); Scott H. Fife, Roy, WA (US); and Christopher H. Rees, Seattle, WA (US)
Assigned to The Boeing Company, Arlington, VA (US)
Filed by The Boeing Company, Chicago, IL (US)
Filed on Oct. 10, 2022, as Appl. No. 18/045,223.
Prior Publication US 2024/0119579 A1, Apr. 11, 2024
Int. Cl. G06T 15/08 (2011.01); G06T 7/00 (2017.01)
CPC G06T 7/0004 (2013.01) [G06T 15/08 (2013.01); G06T 2207/10081 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method for detecting a defect in a powder manufactured component, comprising steps of:
calculating an average value for intensity of each one of a plurality of voxels within a neighborhood of a computed tomography image;
calculating a difference value for the intensity of each one of the voxels in the neighborhood, wherein the difference value is a difference between the average value for the intensity and an original value for the intensity of each one of the voxels;
calculating a standard deviation of the intensity based on the difference value for the voxels in the neighborhood;
calculating a z-score for each one of the voxels;
discarding ones of the voxels in which an absolute value of the z-score is less than a threshold value of the z-score;
identifying a cluster of neighboring ones of the voxels using a clustering algorithm;
determining a cluster-boundary parameter of the cluster; and
classifying the cluster as the defect when the cluster-boundary parameter of the cluster is above a parameter threshold.