A pavement crack detection method based on convolution neural network and image recognition

A convolutional neural network and pavement crack technology, applied in image enhancement, image analysis, image data processing, etc., can solve problems such as large amount of image data, ineffective crack detection, poor pavement crack detection results, etc., and achieve repair The effect of road surface defects, high detection efficiency and fast detection speed

Inactive Publication Date: 2019-01-04
KUNMING UNIV OF SCI & TECH
View PDF2 Cites 26 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Traditional manual detection methods have long been unable to meet the basic needs of road development
At the same time, in the face of large-scale pavement damage images, the traditional image recognition-based methods can no longer meet the application of large-scale data due to the influence of factors such as road image noise, feature extraction method limitations, and large amount of ima

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
  • A pavement crack detection method based on convolution neural network and image recognition
  • A pavement crack detection method based on convolution neural network and image recognition
  • A pavement crack detection method based on convolution neural network and image recognition

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0059] Embodiment 1: a kind of pavement crack detection method based on convolutional neural network and image recognition, the steps are as follows:

[0060] A. Use image acquisition equipment to collect pictures of cracks on the road surface, and preprocess the pictures;

[0061] B, design the convolutional neural network structure and use the processed pictures in step A to train (use various road surface cracks as samples to train the neural network);

[0062] C. Collect road surface pictures and use the trained neural network in step B to judge whether there are defects in the road surface;

[0063] D. If there are defects in the road surface, input the pictures with defects into the trained multi-modal cyclic neural network to obtain the text describing the defects in the road surface, so as to determine the type of cracks.

[0064] Further, the image preprocessing process in the step A includes:

[0065] First, pre-mark the collected pictures, and then pre-process the...

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 a pavement crack detection method based on convolution neural network and image recognition, belonging to the field of traffic pavement detection. The method is based on a convolution neural network chip and a recurrent neural network (the convolution neural network chip is used to detect pavement cracks and the recurrent neural network is used to generate statements describing the types of cracks). The method includes steps: the original image of pavement cracks is pre-marked, the intensity normalization and pixel saturation pretreatment are carried out on the pavement image according to the pre-marking results, the pre-processed pavement images are input into a convolution neural network (CNN) model for training (using different pavements as samples to train theneural network), the network structure and model parameters are determined, the trained network is used to detect the complex pavement, and the type of pavement cracks is determined. The pavement crack detection system with depth learning function is composed of pavement image acquisition equipment, an image pre-processing device, a memory, USB external equipment, a microprocessor and so on.

Description

technical field [0001] The invention relates to a pavement crack detection method based on a convolutional neural network and image recognition, belonging to the field of traffic pavement detection. Background technique [0002] With the rapid development of the highway transportation industry, the maintenance work of the road surface is becoming more and more heavy. The highway management department needs to grasp the information of the road surface quickly and timely. The traditional manual detection method has long been unable to meet the basic needs of road development. At the same time, in the face of large-scale road damage images, traditional image recognition-based methods cannot meet the application of large-scale data due to the influence of factors such as road image noise, feature extraction method limitations, and large image data. [0003] At present, the pavement detection has the following problems: 1. It can only process pavement images under the same condi...

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): G06T7/00
CPCG06T7/0004G06T2207/20081G06T2207/20084
Inventor 朱阳光刘瑞敏王震王枭
Owner KUNMING UNIV OF SCI & TECH
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