Method for automatically detecting and classifying glass defects

Abstract

The invention relates to a method for automatically detecting and classifying glass defects. The method comprises the following steps of: dividing an image of glass by detection software according toa set segmentation threshold value so as to obtain a plurality of divided objects; processing each divided object by the detection software through a neural network classification algorithm, and extracting a plurality of defect feature values corresponding to each divided object; analyzing the defect feature values corresponding to each divided object by the detection software through the neural network classification algorithm, and obtaining a defect type of the divided object. The method for automatically detecting and classifying glass defects is suitable for operations of automatic production of electronic glass and other pieces of glass, so that the glass inspection efficiency is improved, the outgoing quality of the glass is ensured, the labor cost is saved, the automation process ofenterprises is accelerated, benefit is brought to large-scale popularization of automatic glass detection systems, and a wider application range is provided.

Description

technical field [0001] The invention relates to the field of visual image detection, in particular to the field of glass quality detection, and specifically refers to a method for automatic detection and classification of glass defects based on computer software. Background technique [0002] In recent years, with the rapid growth of market demand for glass products, the production of glass products has undergone qualitative changes in terms of quality, variety, and production technology. Especially with the continuous development of production technology, high-end products have higher and higher quality requirements for glass substrates, especially electronic glass, there will be various defects such as bubbles, scratches, and pits on the glass, so the glass is fully guaranteed It is especially important to improve its level of quality. [0003] At present, the defect detection of electronic glass mainly uses manual online detection, which has low precision and high missed...

AI Technical Summary

Problems solved by technology

[0003] At present, the defect detection of electronic glass mainly uses manual online detection, which has low precision and high missed detection rate. Manual detection is easily affected by the subjective factors of the inspectors, and it is easy to cause missed detection of glass defects, especially those with small distortion. Defects are missed, workers are prone to visual fatigue, especially on night shifts, the stability is not high, and the labor cost is high

Images