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

Semantic Segmentation Method of Yellow River Ice Slime Based on Multi-Attention Mechanism and Two-stream Fusion Network

A technology that integrates networks and semantic segmentation. It is applied in neural learning methods, biological neural network models, and computer components. It can solve problems such as poor accuracy and achieve good segmentation results.

Active Publication Date: 2022-05-17
NORTHWESTERN POLYTECHNICAL UNIV
View PDF15 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] In order to overcome the shortcomings of the poor accuracy of existing Yellow River ice detection methods, the present invention provides a semantic segmentation method for Yellow River ice based on a multi-attention mechanism and a two-stream fusion network

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
  • Semantic Segmentation Method of Yellow River Ice Slime Based on Multi-Attention Mechanism and Two-stream Fusion Network
  • Semantic Segmentation Method of Yellow River Ice Slime Based on Multi-Attention Mechanism and Two-stream Fusion Network
  • Semantic Segmentation Method of Yellow River Ice Slime Based on Multi-Attention Mechanism and Two-stream Fusion Network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0018] refer to Figure 1-5 . The specific steps of the Semantic Segmentation Method for the Yellow River Ice Snow Based on the Multi-Attention Mechanism Two-stream Fusion Network of the present invention are as follows:

[0019] 1. Prepare and build the dataset.

[0020] The data sets were taken by different UAVs on the Ningxia-Inner Mongolia section of the Yellow River. After manual selection, the data sets were marked, and the images were marked pixel by pixel. The marks were divided into three categories, including ice, water, and shore. Divide the marked data set into training set, verification set and test set according to the proportion of shooting time. The image size is 1600×640, the training set contains 570 images, the validation set contains 82 images, and the test set contains 244 images.

[0021] 2. Build a semantic segmentation network.

[0022] The network includes two branches, the shallow branch and the deep branch, and in order to better fuse the feature...

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 method for semantic segmentation of Yellow River icicles based on a multi-attention mechanism and dual-stream fusion network, which is used to solve the technical problem of poor accuracy of the existing Yellow River icy detection methods. The technical solution is to firstly collect and mark the data sets, and divide the marked data sets into training data sets and test data sets. Then build a segmentation network structure, the network includes shallow branches and deep branches, add channel attention module in deep branch; add position attention module in shallow branch; fusion module is used for the fusion of shallow branch and deep branch. Put the data in the training set into the network in batches, and use the cross-entropy loss and RMSprop optimizer to train the constructed neural network. Finally, input the image to be tested and use the trained model for testing. The present invention can selectively perform multi-level and multi-scale feature fusion, and capture context information based on the attention mechanism, obtain higher-resolution feature maps, and obtain better segmentation effects.

Description

technical field [0001] The invention relates to a method for detecting Yellow River icicles, in particular to a method for semantic segmentation of Yellow River icicles based on a multi-attention mechanism and dual-stream fusion network. Background technique [0002] Semantic segmentation is a very important field in computer vision. It refers to the recognition of images at the pixel level, that is, to mark the object category to which each pixel in the image belongs, and its goal is to predict the class label of each pixel in the image. River ice monitoring is of great significance to river management in the shipping industry. Accurate ice segmentation is one of the most important techniques in ice regime monitoring research. UAV aerial images have the advantages of high definition, large scale, small area, and high visibility. It can provide prerequisite information for calculating ice sheet density, drift ice velocity, ice sheet distribution, change detection, etc. At...

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): G06V20/10G06V10/26G06V10/764G06V10/82G06K9/62G06N3/04G06N3/08
CPCG06N3/08G06V20/182G06V10/267G06N3/047G06N3/048G06N3/045G06F18/241G06F18/2415
Inventor 张秀伟张艳宁兰泽泽金娇娇余欣范旻昊李春江王亚飞
Owner NORTHWESTERN POLYTECHNICAL UNIV
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