Depth residual error model construction method suitable for hub automatic identification

Abstract

The invention discloses a depth residual error model construction method suitable for hub automatic identification. Firstly, different types of hub images are collected through a camera, a hub image data set is constructed, and the hub data set is divided into a training set and a test set in proportion; a residual block of deep learning is designed, and a hub deep learning model is constructed based on a convolution layer, a pooling layer, a normalization layer, an inactivation layer and a nonlinear activation layer; and the weight is randomly initialized by adopting Xavier, and model parameters are optimized by combining an Adam optimization algorithm and a batch random gradient descent algorithm. The model constructed by adopting the method disclosed by the invention can realize the identification of various hub types, has no over-fitting phenomenon, is high in identification speed and high in identification precision, and can realize the real-time identification of industrial hubs.

Description

technical field [0001] The invention belongs to the field of deep learning, and specifically relates to a method for constructing a deep residual model suitable for automatic wheel hub recognition. Background technique [0002] Under the framework of Industry 4.0, the perception and control technology of smart factories has the characteristics of high speed, high precision, modularization, intelligence, and non-destructive perception. It can realize independent configuration and adaptive adjustment according to different tasks, and meet the needs of customized and personalized products. Adaptive manufacturing, however, traditional perceptual control technology cannot meet the above requirements. Machine vision technology involves many disciplines such as neurobiology, computer science, image processing, pattern recognition and artificial intelligence. It has the characteristics of high efficiency, high precision, non-contact and easy integration, and is the basis for realizi...

AI Technical Summary

Problems solved by technology

On the one hand, due to the huge workload and easy to cause errors in this manual identification method, it is impossible to realize the automatic generation and submission of data that is beneficial to system management

On the other hand, due to the large number of wheel hub types produced by the factory, with the lengthening of the production line, the demand for identification workers is also increasing, which increases the production cost of wheel hub manufacturing, and is a great waste of both manpower and material resources.

Images