Road blocking information extraction based on deep learning image classification

A deep learning and information extraction technology, applied in the fields of instruments, biological neural network models, character and pattern recognition, etc., can solve problems such as difficulty in meeting the urgency of disaster emergency monitoring requirements, multi-image data, and long data processing time. The effect of fitting ability and generalization ability, improving accuracy, and enhancing generalization ability

Active Publication Date: 2019-10-01
AEROSPACE INFORMATION RES INST CAS
View PDF8 Cites 19 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] The method of extracting road blockage information based on pre-disaster and post-disaster multi-temporal images needs to process mor

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 blocking information extraction based on deep learning image classification
  • Road blocking information extraction based on deep learning image classification
  • Road blocking information extraction based on deep learning image classification

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0081] like figure 1 As shown, the technical flow of the CNN-based road blocking information extraction technology of the present application is shown.

[0082] Specifically include the following:

[0083] (1) Using images of typical disaster cases to construct a classification sample library of road blocking images to provide training samples S 0 ;

[0084] (2) Carry out the training of convolutional neural network, obtain initial convolutional neural network model CNNmodel0;

[0085] (3) After obtaining the post-disaster image I(x) and road vector R(x) of the study area x, detect the road blockage and obtain the sample D to be detected n (x);

[0086] (4) Use the trained network CNNmodel0 to treat the detection sample D n (x) Carry out the multi-point road blocking detection that increases detection point along the vertical normal direction of road vector direction:

[0087] (5) Select a small number of road segments to generate new training sample data S 1 , fine-tun...

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 road blocking information extraction based on deep learning image classification, and the method comprises the steps: building a road blocking image classification sample library through employing a disaster typical case image, carrying out the training of a convolutional neural network, and obtaining an initial convolutional neural network model CNNmodel0; obtaining a post-disaster image I (x) and a road vector R (x) of the research area x, and detecting road blocking to obtain a to-be-detected sample Dn (x); using the trained network CNNmodel0 to perform multi-pointroad blocking detection of adding detection points in the vertical normal direction of the road vector direction on the sample Dn (x) to be detected: selecting a small number of road segments to generate new training sample data S1, and performing network fine tuning on the existing network; and repeating the road blocking classification detection work until the detection result meets the precision requirement. Various precision evaluation indexes of the improved convolutional neural network model are superior to those of an original model, and the improved convolutional neural network modelis more suitable for specific problems of disaster area road blocking image classification and detection.

Description

technical field [0001] The invention relates to the technical field of remote sensing monitoring. Specifically, it is road blocking information extraction based on deep learning image classification. Background technique [0002] After a disaster occurs, it is the primary task of emergency rescue to timely and accurately evaluate the road blockage in the disaster area, grasp the distribution and quantity of damaged roads, and the traffic situation of the main roads in the disaster area. The emergency assessment of post-disaster loss is to quickly assess the scale of the disaster and the degree of disaster damage and loss immediately after the disaster. The assessment of the degree of road blockage in the disaster area is an important part of the emergency assessment of post-disaster loss. Road blockage information in the disaster area serves the real-time reporting of the disaster situation and the formulation of disaster relief decisions, and provides a scientific basis fo...

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): G06K9/00G06K9/62G06N3/04
CPCG06V20/182G06N3/045G06F18/214
Inventor 王世新王福涛杨宝林周艺
Owner AEROSPACE INFORMATION RES INST CAS
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