CNN-deep-learning-based DBCC classification model and constructing method thereof

A classification model and deep learning technology, applied in image analysis, character and pattern recognition, image data processing, etc., can solve problems such as inaccurate recognition of small pictures and poor accuracy of crack recognition

Inactive Publication Date: 2017-06-30
SHAANXI NORMAL UNIV
View PDF3 Cites 37 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to overcome the problems of inaccurate identification of small pictures with 16*16 pixel resolution and poor accuracy of crack identification by the CIFAR10 model

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
  • CNN-deep-learning-based DBCC classification model and constructing method thereof
  • CNN-deep-learning-based DBCC classification model and constructing method thereof
  • CNN-deep-learning-based DBCC classification model and constructing method thereof

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0035] refer to Figure 1-Figure 4 , a method for building a DBCC classification model based on CNN deep learning, comprising the following steps:

[0036] (1) Convolute and sum the input original picture with all the convolution kernels in the first convolutional layer in a convolutional manner to obtain the feature map of the first convolutional layer;

[0037] (2) Add a Relu activation function after the first convolutional layer;

[0038] (3) After the first convolutional layer, add a local response value normalization layer for picture brightness correction, and the local response value normalization layer improves the recognition effect of the network;

[0039] (4) Downsample the feature map of the first convolutional layer in the first pooling layer, reduce the resolution and select excellent features as the feature map of the first pooling layer;

[0040] (5) On the second convolutional layer, the feature map of the first pooling layer and all the convolution kernels...

Embodiment 2

[0076] refer to Figure 5 , describe in detail the operation process of bridge crack detection and location:

[0077] The first step is to use image acquisition equipment to collect five bridge crack pictures with different background textures and materials. The total number of collected pictures is 2000, and all pictures are normalized into 1024*1024 resolution pictures. The pictures are divided into 2 data sets, artificially amplified data set and test data set, each data set has 1000 pictures;

[0078] In the second step, use a W*H fixed-size window to slide non-overlapping 1000 pictures in the artificial amplification data set, and at the same time, use the small slice of the bridge crack picture covered by the sliding window as a ROI region of interest. Among them, the small slice image containing the bridge background is called the bridge background surfel, and the small slice containing the bridge crack is called the bridge crack surfel. The specific process is shown i...

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 discloses a CNN-deep-learning-based DBCC classification model and the constructing method thereof. The DBCC classification model consists of four convolution layers, three pooling layers and two full-connection layers. The DBCC classification model uses the softmax loss function as the loss function; the activation function (RELU) is added behind the first convolution layer, the fourth convolution layer, the second pooling layer, the third pooling layer, the first full-connection layer respectively; and a local response normalized layer LRN is added behind the first convolution layer, and a dropout layer is added behind the first full-connection layer. The DBCC classification model of the present invention is constructed based on a convolution neural network CNN, which increases the depth of the network through the use of more convolution cores in each convolution layer, the addition of the LRN and the use of the dropout so that the DBCC classification model achieves higher identification accuracy in its identification of small pictures with 16 * 16pixel resolutions.

Description

technical field [0001] The invention belongs to the field of image processing and computer vision, and specifically relates to a CNN deep learning-based DBCC classification model and a construction method. Background technique [0002] As the hub of traffic systems such as roads, highways, and railways, bridges need to be regularly evaluated for their health status, and bridge cracks, as one of the most important bridge diseases, seriously affect the safe operation of bridges, and more serious ones will occur Bridge crash accident. Therefore, it is very important to effectively detect and identify bridge cracks. [0003] Most of the current research on bridge crack detection is based on image processing algorithms. The core of the image processing algorithm is the window sliding algorithm, constructing the training set and using the training set to train the classification model. Specifically, after constructing the training set by using the window sliding algorithm and t...

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
Patent Type & Authority Applications(China)
IPC IPC(8): G06T7/00G06T7/62G06K9/62
CPCG06T7/0004G06T2207/20084G06T2207/20081G06T2207/30132G06T2207/20104G06F18/2415
Inventor 李良福马卫飞李丽张玉霞
Owner SHAANXI NORMAL UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap