| CPC G01S 17/89 (2013.01) [G06N 3/08 (2013.01)] | 20 Claims |

|
1. A method, comprising:
training a neural network to identify a deformity associated with an object;
aligning, by a processing device, a measurement point cloud for an object with reference data for the object;
determining, by the processing device using the neural network, a deviation associated with the object;
wherein the measurement point cloud is a three-dimensional (3D) point cloud;
transforming, by the processing device, the 3D point cloud into a 2.5-dimensional (2.5D) matrix; and
performing evaluations of the 2.5D matrix using the neural network on-the-fly during scanning of the object as the points of the 3D point cloud are received and processed by the processing device to identify the deviation associated with the object based on how the object is normally without the deviation.
|