Title:
MACHINE VISION-BASED MACHINE TOOL PART ONLINE INSPECTION METHOD
Document Type and Number:
WIPO Patent Application WO/2023/134793
Kind Code:
A3
Abstract:
The present invention relates to the technical field of machine vision, and relates in particular to a machine vision-based machine tool part online inspection method. The method comprises: acquiring confidence degrees that first pixel points in an image to be inspected of a bevel gear belong to a scratch category; determining a first pixel point with a maximum confidence degree in a semantic fuzzy region as an initial seed point, and performing region growth in the semantic fuzzy region; obtaining a similarity between a second pixel point and the seed point in a set growth window in the semantic fuzzy region; during a region generation process, determining category labels of second pixel points; obtaining a probability index that a second pixel point having a fuzzy label belongs to the scratch category, and updating the category label of the fuzzy label according to the probability index; and determining first pixel points in a scratch region and second pixel points, the category labels of which are the scratch category, as scratch pixel points, and inspecting scratches in the bevel gear according to the scratch pixel points. The present invention increases the accuracy of scratch inspection of a bevel gear in a machine tool part.
Inventors:
ZHOU HAO (CN)
Application Number:
PCT/CN2023/086012
Publication Date:
September 07, 2023
Filing Date:
April 03, 2023
Export Citation:
Assignee:
SUZHOU MAICHUANG INFORMATION TECH CO LTD (CN)
International Classes:
G06T7/00; G06T7/11; G06V10/26
Foreign References:
| CN115601361A | 2023-01-13 | |||
| CN113689428A | 2021-11-23 | |||
| US20210110174A1 | 2021-04-15 |
Attorney, Agent or Firm:
SUZHOU ZHOUZHI PATENT AGENCY (SPECIAL GENERAL PARTNERSHIP) (CN)
Download PDF: