Concrete crack recognition method based on convolutional neural network

Abstract

The invention discloses a concrete crack recognition method based on a convolutional neural network. The method comprises the steps of acquiring a concrete crack picture and cutting the picture into apreset size; rotating, cutting and transforming the picture after size cutting to obtain a picture expansion data set; performing gray processing on the pictures in the picture expansion data set toobtain a to-be-used image data set; constructing a convolutional neural network, and dividing the to-be-used image data set into a training set, a test set and a verification set according to a presetproportion; training the constructed convolutional neural network, storing the trained convolutional neural network, and selecting and storing a convolutional neural network model with the highest classification accuracy by utilizing the verification set; and inputting the test set into the selected convolutional neural network, and outputting an recognition result. According to the method, the concrete crack picture is recognized by utilizing the feature extraction capability of the convolutional neural network, the recognition accuracy is improved compared with a traditional recognition method, and meanwhile, the recognition efficiency is improved based on the rapid calculation capability of the convolutional neural network.

Description

technical field [0001] The present invention relates to the field of image recognition, and more specifically, to a concrete crack recognition method based on a convolutional neural network. Background technique [0002] Concrete structure is the most widely used structural form in civil engineering, and health monitoring of in-service concrete structures has become an important research field. The most intuitive information to characterize structural damage is the appearance and expansion of cracks. As an important characteristic phenomenon of concrete structure damage, cracks have become an important object for studying the health status of concrete structures. [0003] Cracks are universal in concrete structures, and their occurrence and development have certain regularity, but the development of cracks in concrete structures is not stable enough, which is manifested in: (1) The direction of cracks is irregular, and bifurcations appear during the development of cracks (2...

AI Technical Summary

Problems solved by technology

[0004] In order to overcome the defects of low concrete crack recognition efficiency and poor accuracy in the above-mentioned prior art, the present invention provides a concrete crack recognition method based on convolutional neural network

Images