Building wood crack identification method based on convolutional neural network

The invention is suitable for the interdisciplinary field of deep learning and civil engineering, and relates to a building wood crack recognition method based on a convolutional neural network, and the method comprises the steps: collecting a building wood crack image, and selecting a crack-contain...

Full description

Saved in:
Bibliographic Details
Main Authors PANG GAOZHAO, LEI JIANWEI, FANG HONGYUAN, XUE BINGHAN, DONG JIAXIU, MA DUO, WANG NIANNIAN
Format Patent
LanguageChinese
English
Published 22.01.2021
Subjects
CALCULATING
COMPUTING
COUNTING
IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
PHYSICS
Online AccessGet full text

Cover

More Information
Summary:The invention is suitable for the interdisciplinary field of deep learning and civil engineering, and relates to a building wood crack recognition method based on a convolutional neural network, and the method comprises the steps: collecting a building wood crack image, and selecting a crack-containing picture; marking building wood crack data sets in a classified mode, and creating and dividing abig data set and a small data set into a training set, a verification set and a test set in proportion; carrying and initializing the convolutional neural network model, and training the model basedon the big data set and the small data set; setting different hyper-parameters, introducing a verification set picture test, and searching an optimal hyper-parameter value; testing the model accordingto the obtained optimal model, outputting various detection numerical value indexes, and judging whether an expected value is reached or not; and comparing and analyzing an existing image processingmethod. The building wood cr
Bibliography:Application Number: CN202011203806