Unlock instant, AI-driven research and patent intelligence for your innovation.

Road surface damage identification method based on improved Faster R-CNN

A recognition method, broken technology, applied in the direction of neural learning methods, character and pattern recognition, instruments, etc., can solve the problem of expensive convolution

Inactive Publication Date: 2021-07-30
HOHAI UNIV
View PDF0 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

But since it is applied in high-resolution feature maps, this type of convolution is very expensive in practice

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
  • Road surface damage identification method based on improved Faster R-CNN
  • Road surface damage identification method based on improved Faster R-CNN
  • Road surface damage identification method based on improved Faster R-CNN

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0035] The present invention will be further described below. The following examples are only used to illustrate the technical solution of the present invention more clearly, but not to limit the protection scope of the present invention.

[0036] Such as figure 1 As shown, the present invention provides a road damage recognition method based on the improved Faster R-CNN, which includes three parts: data set classification and labeling, model training and output classification results. Specific steps are as follows:

[0037] Step 1: Collect road damage pictures and smooth road pictures from various websites, perform data enhancement operations, and classify and mark road damage pictures according to the degree of danger and pothole size. The specific steps are as follows:

[0038] Step 11: Collect damaged pavement pictures and smooth pavement pictures in the network, 70% of which include cracks, small potholes, medium potholes and large potholes, as positive samples; the rem...

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 provides a pavement damage identification method based on improved Faster R-CNN, which applies a data enhancement method, classifies pavement damage images in advance, extracts pavement damage features by improving a VGG16 convolutional layer, refines the granularity of a feature map, reduces the influence caused by background factors, introduce a Sigmoid weighting function to improve a non-maximum suppression algorithm, solves the problem of pothole leak detection, and quickly positions potholes in an image by using RPN and are counted. According to the method, the problems of low pavement damage recognition rate and missing detection and error detection during detection are solved.

Description

technical field [0001] The invention relates to the technical field of road damage recognition, in particular to a road damage recognition method based on improved Faster R-CNN. Background technique [0002] In recent decades, with the rapid development of my country's economy, great achievements have been made in infrastructure construction, and among them, the achievements of road construction are eye-catching. According to statistics from the Ministry of Transport, since 2010, the total mileage of national highways has grown rapidly. By the end of 2019, the total mileage of national highways has reached 5.0125 million kilometers, and the road density has reached 52.31 kilometers per 100 square kilometers. However, in actual use, the pavement is often affected by weather factors (such as excessive sunlight, sudden temperature changes, high humidity, excessive wind, etc.) and traffic loads, resulting in a reduction in the life of the pavement, causing traffic delays, and af...

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): G06K9/62G06K9/46G06N3/04G06N3/08
CPCG06N3/084G06V10/40G06N3/045G06F18/241
Inventor 平萍刘元佳吕鑫毛莺池许国艳
Owner HOHAI UNIV