	US 12,456,185 B2
	Automated machine vision-based defect detection
Rajen Bhatt, McDonald, PA (US); Shitong Mao, Pittsburgh, PA (US); Raviprakash Kandury, Cupertino, CA (US); Michelle Tai, Fremont, CA (US); and Geoffrey Newman, Pittsburgh, PA (US)
Assigned to Qeexo, Co., San Jose, CA (US)
Filed by QEEXO, CO., San Jose, CA (US)
Filed on Nov. 16, 2023, as Appl. No. 18/511,909.
Application 18/511,909 is a continuation of application No. 18/064,040, filed on Dec. 9, 2022, granted, now 11,847,775.
Application 18/064,040 is a continuation of application No. 17/110,131, filed on Dec. 2, 2020, granted, now 11,538,146, issued on Dec. 27, 2022.
Claims priority of provisional application 62/950,440, filed on Dec. 19, 2019.
Prior Publication US 2025/0117917 A1, Apr. 10, 2025
This patent is subject to a terminal disclaimer.
Int. Cl. G06T 7/00 (2017.01); G06N 3/08 (2023.01); G06T 7/11 (2017.01); G06T 7/194 (2017.01)

CPC G06T 7/0004 (2013.01) [G06N 3/08 (2013.01); G06T 7/11 (2017.01); G06T 7/194 (2017.01); G06T 2207/20076 (2013.01); G06T 2207/20081 (2013.01); G06T 2207/20084 (2013.01); G06T 2207/30164 (2013.01)]

20 Claims

1. A method comprising:

obtaining a test image of an object;

pre-processing the test image to generate a plurality of regions of the test image for subsequent processing steps;

segmenting portions of the test image into a plurality of patches for input into a neural network that was trained using a historical dataset that does not include the test image, each patch of the plurality of patches corresponding to a segmented portion of the test image;

determining whether each patch of the plurality of patches is within a region of the plurality of regions of the test image; and

inputting each patch of the plurality of patches that is determined to be within a region of the plurality of regions into the neural network as a respective matrix representation, while excluding patches of the plurality of patches that are determined to not be within the region of the plurality of regions from the neural network.