Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

A Road Extraction Method Based on Prediction and Residual Refinement Network

A predictive network and road extraction technology, applied in neural learning methods, biological neural network models, character and pattern recognition, etc., to achieve the effect of improving clarity, reducing information loss, and reducing information loss

Active Publication Date: 2022-07-05
HUBEI UNIV OF TECH
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Due to the variable road types in aerial images or high-resolution remote sensing images, road extraction still faces severe challenges

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 Road Extraction Method Based on Prediction and Residual Refinement Network
  • A Road Extraction Method Based on Prediction and Residual Refinement Network
  • A Road Extraction Method Based on Prediction and Residual Refinement Network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0019] In order to facilitate the understanding and implementation of the present invention by those skilled in the art, the present invention will be described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the embodiments described herein are only used to illustrate and explain the present invention, but not to limit it. this invention.

[0020] see figure 1 , the present invention provides a kind of road extraction method based on prediction and residual refinement network, comprising the following steps:

[0021] Step 1: Design the aerial image road extraction network combining the prediction network and the residual refinement network;

[0022] In this embodiment, the prediction network adopts the Encoder-Decoder structure, combined with the hole convolution module DCM and the multi-layer pooling module MPM;

[0023] Recently, the U-Net algorithm has achieved very good results in the field of cell segmen...

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 extraction method based on prediction and residual refinement network, and designs an aerial image road extraction algorithm combining prediction network and residual refinement network according to different characteristics of road targets. The prediction network adopts the Encoder-Decoder structure; secondly, the prediction network combines the atrous convolution module DCM and the multi-core pooling module MPM, which can fully obtain the context information and road edge information of the image and improve the road edge detection ability. The residual refinement network will refine the initial road prediction results generated by the prediction network and improve the road fuzzification caused by noise in the prediction network. The network also integrates BCE, SSIM and IoU loss functions for supervised training to reduce road information loss. It is beneficial to extract the complete road structure.

Description

technical field [0001] The invention belongs to the technical fields of digital image processing, pattern recognition and machine learning, and relates to a road extraction method, in particular to a road extraction method based on prediction and residual refinement network. Background technique [0002] Road extraction can be formulated as a binary classification problem, similar to multi-object segmentation in semantic segmentation, where road segmentation reduces the segmentation categories in comparison. Due to the variable road types in aerial images or high-resolution remote sensing images, road extraction still faces severe challenges. After years of research, a large number of algorithms have emerged for road extraction from aerial images, which can be roughly divided into three categories: feature-based, object-based, and knowledge-based road extraction methods. The road extraction algorithm based on feature level includes template matching method, edge and paralle...

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 Patents(China)
IPC IPC(8): G06V20/56G06V20/58G06N3/08
CPCG06N3/08G06V20/588G06V20/56Y02T10/40
Inventor 熊炜管来福李敏王娟李利荣曾春艳刘敏
Owner HUBEI UNIV OF TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com