公开(公告)号：US20240071059A1

公开(公告)日：2024-02-29

申请号：US18269657

申请日：2021-12-23

Applicant: 3M INNOVATIVE PROPERTIES COMPANY

Inventor： Samuel S. Schreiner ,  Steven P. Floeder ,  Jeffrey P. Adolf ,  Carl J. Skeps ,  Shane T. Van Kampen

IPC: G06V10/778 ,  G06V10/77

CPC classification number: G06V10/778 ,  G06V10/7715

Abstract: An example method for selecting product images for training a machine-learning model includes obtaining product images to include in an image population; receiving an indication of an image selection strategy for determining if a product image is to be included in a set of images of interest; determining image transforms based on configuration data for the indicated image selection strategy, wherein the image transforms perform image manipulation operations to obtain transformed image data for each of the product images in the image population; selecting a subset of images from the image population for inclusion in the set of images of interest based on the indicated image selection strategy and the transformed image data; determining one or more descriptive labels and applying the one or more descriptive labels to the respective sets of images; and training an inspection model for a product inspection system based on the labeled images.

公开(公告)号：US20240046617A1

公开(公告)日：2024-02-08

申请号：US18268834

申请日：2021-12-23

Applicant: 3M INNOVATIVE PROPERTIES COMPANY

Inventor： Jeffrey P. Adolf ,  Steven P. Floeder ,  Samuel S. Schreiner ,  Scott P. Daniels

IPC: G06V10/764 ,  G06T7/00 ,  G06V10/772 ,  G06V10/774 ,  G06V10/82

CPC classification number: G06V10/765 ,  G06T7/0004 ,  G06V10/772 ,  G06V10/774 ,  G06V10/82 ,  G06T2207/20081 ,  G06T2207/30108

Abstract: Machine learning-based systems are described for automatically generating an inspection recipe that can be utilized when inspecting web materials, sheet parts or other products for defects. One example method for generating an inspection recipe includes assigning, by processing circuitry, a pseudo-label to each image of a plurality of images based on the content of the image and a labeling model, and storing the pseudo-labeled images in a second set of images; extracting, by the processing circuitry, one or more features from each image of the plurality of images, and storing the one or more features as a feature list for the image; generating, by the processing circuitry, a decision tree based on the second set of images and the feature lists of the second set of images; generating, by the processing circuitry, an inspection recipe based on the decision tree, the inspection recipe comprising a plurality of classification rules; and outputting the inspection recipe.