Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Automatic recognition and intelligent positioning method for seismic damages to reinforced concrete structure based on computer vision

A reinforced concrete, computer vision technology, applied in computer parts, calculation, character and pattern recognition, etc., to meet the needs of online monitoring and early warning and real-time data processing, improve automation, improve efficiency and the effect of

Pending Publication Date: 2018-09-11
HARBIN INST OF TECH
View PDF3 Cites 7 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to solve the shortcomings of existing reinforced concrete structure earthquake damage identification methods that need to provide camera internal and external parameters for image shooting or require additional professional measuring equipment, and propose an automatic identification of reinforced concrete structure earthquake damage based on computer vision with smart positioning methods

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Automatic recognition and intelligent positioning method for seismic damages to reinforced concrete structure based on computer vision
  • Automatic recognition and intelligent positioning method for seismic damages to reinforced concrete structure based on computer vision
  • Automatic recognition and intelligent positioning method for seismic damages to reinforced concrete structure based on computer vision

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment approach 1

[0017] Specific implementation mode one: the method for automatic identification and intelligent positioning of reinforced concrete structure earthquake damage based on computer vision in this embodiment, such as figure 1 shown, including:

[0018] Step 1. Downsample the input image, manually mark the damaged area of ​​the downsampled image with a rectangular frame according to the preset damage type, obtain data representing the position and size of the rectangular frame, and mark the damage area according to the damage type Label the damaged area.

[0019] For example, in one embodiment, a color image may be down-sampled to 640x640x3 to reduce computational cost. MATLAB parallel program imageLabeler can be used to generate rectangular box labels.

[0020] There are many choices for the data representing the position and size of the rectangular box, for example, you can use the horizontal / vertical coordinates of the pixel of the upper left corner of the rectangle and the le...

specific Embodiment approach 2

[0028] Specific implementation mode two: the difference between this implementation mode and specific implementation mode one is that step one specifically includes:

[0029] Step 11. Down-sample the input image, set 4 types of damage, namely concrete cracking, concrete spalling, steel bar exposure, and steel bar buckling, and use a rectangular frame to map the damage area in the down-sampled image according to the 4 damage types Perform manual marking to obtain the coordinates of the upper left corner of the rectangular frame and the pixel values ​​of length and width, and label the damaged area according to the type of damage.

[0030] Step 1 and 2: Rotate the input image counterclockwise by 90 degrees, 180 degrees, 270 degrees, flip it horizontally, and flip it vertically to obtain rotated or flipped images respectively, and process the obtained images in steps one by one.

[0031] The beneficial effect of this embodiment is that the training samples are expanded, which can...

specific Embodiment approach 3

[0033] Specific embodiment three: the difference between this embodiment and specific embodiment one or two is: in step two, the structure of each layer of deep neural network is:

[0034] L0 layer: The input has a width of 32 and a depth of 3. Perform the convolutional layer operation, the width of the convolutional layer operation is 7, the depth is 3, the number is 16, the step is 1, and the zero padding is 0.

[0035] L1 layer: The input has a width of 26 and a depth of 16. Perform activation layer operations.

[0036] L2 layer: The input width is 26, the depth is 16, and the regularization layer operation is performed.

[0037] L3 layer: The input has a width of 26 and a depth of 16. Perform the convolutional layer operation. The width of the convolutional layer operation is 5, the depth is 16, the number is 32, the step is 1, and the zero padding is 0.

[0038] L4 layer: The input has a width of 22 and a depth of 32. Perform activation layer operations.

[0039] L5 l...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention relates to an automatic recognition and intelligent positioning method for seismic damages to a reinforced concrete structure based on computer vision, and aims at ironing out the defects of a conventional seismic damage recognition method for the reinforced concrete structure needs the interior and exterior parameters of an imaging camera or needs an additional professional measurement equipment and a corresponding intelligent algorithm. The method comprises the steps: carrying out the downsampling of an input image, marking a damaged region of the downsampled image according tothe preset damage type through a rectangular frame, obtaining data for indicating the position and size of the rectangular frame, and marking the damaged area according to the damage type; inputtinga training set into a deep convolutional neural network for training, wherein a loss function used in the training process is a multi-objective optimization function, and an optimization algorithm isa stochastic gradient descent algorithm with momentum; inputting a resampled to-be-recognized image into the trained deep convolutional neural network, and obtaining a recognition result. The method is used for the health monitoring and disaster prevention and reduction for the civil engineering.

Description

technical field [0001] The invention relates to the fields of civil engineering health monitoring and disaster prevention and mitigation, in particular to a computer vision-based automatic earthquake damage identification and intelligent positioning method for reinforced concrete structures. Background technique [0002] With the rapid development of my country's national economic construction, more and more reinforced concrete structures play an extremely important role. Under the action of earthquake load, reinforced concrete structures will often form different degrees of local damage, such as concrete cracking, concrete spalling, steel bar exposure, steel bar buckling, etc., which will affect the service safety of the structure, resulting in attenuation of resistance or serviceability of the structure. fall, which may lead to catastrophic accidents in extreme cases. Therefore, after the structure has experienced an earthquake, the management department will invest a lot...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): G06K9/62G06N3/04
CPCG06N3/045G06F18/24G06F18/214
Inventor 李惠徐阳鲍跃全
Owner HARBIN INST OF TECH
Features
  • Generate Ideas
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com