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

Convolutional neural network road damage identification method based on extended Kalman filtering

A convolutional neural network and extended Kalman technology, applied in the field of image processing, can solve problems such as classification accuracy and high computational complexity, and achieve the effects of reducing calculation and recognition time, improving accuracy, and reducing dimensions

Inactive Publication Date: 2019-08-20
HANGZHOU DIANZI UNIV
View PDF2 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the super-high classification accuracy of deep learning convolutional neural networks is at the cost of super-high computational complexity.

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
  • Convolutional neural network road damage identification method based on extended Kalman filtering
  • Convolutional neural network road damage identification method based on extended Kalman filtering
  • Convolutional neural network road damage identification method based on extended Kalman filtering

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0056] The present invention will be further described below.

[0057] Such as figure 1 As shown, the road damage recognition method based on the convolutional neural network of the extended Kalman filter, the specific steps are as follows:

[0058] Step 1. Image preprocessing.

[0059] 1. Step 1. There are n road damage images containing damage and each image has a certain proportion of information loss. For example, the information loss rate of the sth image is 1%, 5%, 10%, 15% and 20%, and sort the images of each scale; set the resolution of the i-th damage image as v i × h i , v i is the number of pixels in a row on the i-th damage map; h i is the number of pixels in a column on the i-th damage map, i=1,2,...,n; the damage category of the i-th damage map is z i ;

[0060] Step 2, image enhancement and convolutional neural network training.

[0061] 2.1, i=1, 2, ..., n, execute steps 2.2 to 2.4 in sequence.

[0062] 2.2. Enlarge the i-th damage map obtained in step...

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 convolutional neural network road damage identification method based on extended Kalman filtering. Before the road is repaired, a road maintainer needs to research the road surface condition, and the project needs to consume huge manpower, material resources and financial resources. The method comprises the following steps: 1, preprocessing an image; and 2, carrying out image enhancement and convolutional neural network training; 3, k = 1, 2,..., m, and executing the step 4 to the step 6 in sequence, wherein m is the number of the detected images; 4, amplifying the kth tested image and adjusting the kth tested image to 300 * 300 resolution; 5, inputting the kth tested extended image obtained in the step 5 into the convolutional neural network obtained by trainingin the step 2; and 6, optimizing the weight initial value obtained in the step 5 through an extended Kalman filtering algorithm. According to the method, real-time parameter updating is carried out byadopting methods such as feedforward operation, a random gradient descent method, feedback operation, PCA dimension reduction and extended Kalman filtering, and a convolutional neural network model with high accuracy is established.

Description

technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to a road damage recognition method based on a convolutional neural network of an extended Kalman filter. Background technique [0002] Cracks in damaged roads are the main problem faced by road maintenance today, and the restoration of many damaged roads is a huge project. Before repairing, road maintenance workers need to conduct research on the road surface conditions, and this project requires huge manpower, material and financial resources. With the development of high technology, scientists have thought of a simple method, using sophisticated camera equipment placed on the front of the car, can get a lot of useful pictures in a short time, and then filter out useful road damage images. [0003] The collected road damage images are divided into 8 categories, and the classification is as follows: D00 indicates the longitudinal wheel mark part of the linear ...

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): G06K9/00G06K9/62
CPCG06V20/588G06F18/214
Inventor 李艳文成林
Owner HANGZHOU DIANZI UNIV