Reinforcing steel bar model training method and device based on convolutional neural network

Abstract

The embodiment of the invention provides a reinforcing steel bar model training method and device based on a convolutional neural network, and the method comprises the steps: carrying out the sampling of an original image of the end face of a reinforcing steel bar, obtaining a training sample, and training the training sample; performing down-sampling on a convolutional layer in the convolutional neural network model to obtain a first detection channel for predicting a large-size target; performing up-sampling on output in the first detection channel, and performing image fusion on an up-sampling result and a feature image in the convolutional layer to obtain a second detection channel for predicting a medium-sized target and a third detection channel for predicting a small-sized target; and deleting the first detection channel of the convolutional layer to obtain an updated convolutional neural network model, and detecting the training sample through the updated convolutional neural network model to obtain the quantity information of the reinforcing steel bars. By adopting the method, the training parameters can be reduced when the convolutional neural network model is used for training, namely, the time and memory required for training the model are reduced.

Description

technical field [0001] The invention relates to the technical field of visual recognition, in particular to a method and device for training a steel bar model based on a convolutional neural network. Background technique [0002] In various constructions, such as urban infrastructure, roads, bridges, etc., or civil buildings, etc., steel bars are required as basic materials. Statistics show that the annual output of steel bars in China has exceeded 200 million tons, which is an indispensable material in the construction industry. [0003] Generally, when steel bars are used as basic materials, human resources are required to count the number of steel bars entering the construction site. However, it is time-consuming and the accuracy cannot be guaranteed. Therefore, there are some methods of counting steel bars through convolutional neural networks. method. [0004] However, the current method of counting steel bars through the convolutional neural network only saves human ...

AI Technical Summary

Problems solved by technology

[0003] Generally, when steel bars are used as basic materials, human resources are required to count the number of steel bars entering the construction site. However, it is time-consuming and the accuracy cannot be guaranteed. Therefore, there are some methods of counting steel bars through convolutional neural networks. method

[0004] However, the current method of counting steel bars through the convolutional neural network only saves human resources, because it takes a long time to collect samples and model training to obtain the results of steel bar counting, and also requires a large amount of memory for corresponding data processing. May affect the progress of other steps of the project

Images