Neural Networks As A Tool For Pattern Recognition of Fasteners

Abstract

The work is devoted to the study of pattern recognition features of industrial parts in individual fasteners' forms. The main types of neural network architectures and their features are considered. Neural networks are classified into separate categories for ease of perception and analysis. An approach to recognition of hardware products such as fasteners using neural network, which is implemented in Python using Keras machine learning library, is proposed. The main generators are described: for training data, testing, and validation. Codes fragments of corresponding programs for implementation of the proposed approach to pattern recognition of fasteners are presented