Optical fiber end face defect detection method, equipment and storage medium

Abstract

The invention relates to an optical fiber end face defect detection method, equipment and storage medium, and specifically includes the following steps: collecting a normal sample data set and a defect sample data set of an optical fiber end face, constructing and training a conditional generation confrontation network, and capable of performing defect sample data expansion, It solves the problem of insufficient sample data for defect detection and avoids the occurrence of overfitting; builds and trains a cyclic generative adversarial network, uses the generator to detect defects on the fiber end face, and also generates new defect samples to further expand defects Sample data set; according to the expanded defect sample set and normal sample data set, the cyclic generative adversarial network model after adversarial training is used to identify defects, and the defect area is obtained. This method belongs to the unsupervised learning method, which provides a brand-new method for defect detection, which not only saves the trouble of marking pictures, but also greatly improves the accuracy of defect detection.

Description

technical field [0001] The invention belongs to the technical field of artificial intelligence, and in particular relates to an optical fiber end face defect detection method, equipment and storage medium. Background technique [0002] The current defect detection methods mainly include artificial visual method, defect detection method based on machine vision and defect detection method based on deep learning. The defect detection method based on deep learning is the current mainstream method. Deep learning can fit complex functions by training and learn a certain deep neural network, characterize the distributed representation of input data, and has a strong ability to automatically learn the essential characteristics of data from limited samples. Ability. Although there are a large number of excellent deep learning algorithms, these algorithms often need a lot of data-driven to maximize their performance. In the actual production process, there are problems such as fewer...

AI Technical Summary

Problems solved by technology

[0003] In order to solve the technical problems of low detection efficiency and poor detection effect of the defect detection methods commonly used in the prior art, the present invention provides a method for detecting defects on optical fiber end faces, which mainly includes:

Images