Machine vision defect detection method and platform based on edge computing

Abstract

The invention provides a machine vision defect detection method based on edge calculation, an edge calculation defect detection platform and a machine vision platform. The method comprises the steps:enabling the edge computing defect detection platform to receive an equipment image sent by the machine vision platform; establishing an equipment model based on the equipment image; enabling the machine vision platform to perform defect detection on the equipment on the production line by utilizing the equipment model and obtaining a missing detection probability; enabling the edge calculation defect detection platform to evaluate the maturity of the equipment model through the missed detection probability to continuously perfect the iterative repair equipment model; until the equipment modelwith the maturity meeting the preset requirement is obtained, and issuing the equipment model to the cloud application platform. According to the invention, the method has the advantages that the method is good in adaptability to to-be-detected products when being used for defect detection, good in detection effect on deformation defects of most products under the condition of comprehensively considering indexes such as missing detection probability, high in accuracy and capable of remarkably shortening detection time and guaranteeing stable and reliable operation, and efficiency of the method is far higher than that of manual operation.

Description

technical field [0001] The present disclosure relates to the technical field of defect detection, in particular to an edge computing-based machine vision defect detection method, an edge computing defect detection platform, and a machine vision platform. Background technique [0002] At present, most of the defect detection of industrial equipment relies on the visual inspection of engineers or years of work experience for defect detection. Due to the limited ability of engineers during the detection process, visual inspection or empirical detection cannot achieve high accuracy, which makes it difficult for engineers to detect equipment defects. , it is impossible to comprehensively detect equipment defects, which will easily lead to production loss and property waste caused by incomplete equipment detection. [0003] In view of the defects of manual inspection equipment, it is no longer able to meet the market demand, and it is replaced by machine vision inspection. Machin...

AI Technical Summary

Problems solved by technology

[0004] However, the existing machine vision system has poor adaptability, including poor adaptability to the different characteristics of different types of products (differences in components, defect categories and states, etc.), poor adaptability to the external environment of equipment operation, equipment debugging and The operability is poor. In the case of comprehensive consideration of indicators such as the probability of missed detection, the detection effect of most products with deformation defects is not ideal, and the accuracy rate is low.

Images