Automatic Identification Technology of Optical Fiber based on Genetic Neural Network Algorithm

Abstract

With the development of image recognition technology, many manufacturers design and manufacture different fiber cleaning machines. Using the advantages of neural network(NN) in pattern recognition, this paper proposes a genetic neural network algorithm(NNA) combining NN and genetic algorithm to study the automatic optical fiber recognition technology. Optical fiber image recognition and image preprocessing are discussed, and the hardware implementation of optical fiber automatic recognition system is studied; The Gabor extracted image texture features are analyzed, and the simulation test of optical fiber image recognition is carried out according to the Gabor extracted feature training. The test results verify the accuracy of the optical fiber automatic recognition technology proposed in this paper.