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

Deep learning and multi-source remote sensing data-based ocean anomaly mesoscale eddy identification method

A recognition method and deep learning technology, applied in neural learning methods, character and pattern recognition, image data processing, etc., can solve problems such as unclear generation mechanism and insufficient understanding of spatio-temporal distribution characteristics

Active Publication Date: 2020-12-18
INST OF OCEANOLOGY - CHINESE ACAD OF SCI
View PDF3 Cites 11 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, there is currently a lack of systematic research on anomalous mesoscale eddies, and people's understanding of the temporal and spatial distribution characteristics of anomalous mesoscale eddies in the global ocean is still not clear enough, and its formation mechanism is still unclear

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
  • Deep learning and multi-source remote sensing data-based ocean anomaly mesoscale eddy identification method
  • Deep learning and multi-source remote sensing data-based ocean anomaly mesoscale eddy identification method
  • Deep learning and multi-source remote sensing data-based ocean anomaly mesoscale eddy identification method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0035] A method for identifying marine anomalous mesoscale eddies based on deep learning and multi-source remote sensing data, the method includes the following steps, such as figure 1 Shown:

[0036] S1. Construction of a mesoscale anomalous vortex sample library:

[0037] ①Extract mesoscale eddies: extract mesoscale vortices based on SSHA data; first perform high-pass filtering on the global SSHA with a radius of 5° (latitudinal) and 10° (longitudinal) to remove large-scale signals, and then conduct SSHA contours Search and filter. If the closed SSHA contour line meets the following conditions: a) There is only one extreme point in the contour line, and the extreme point refers to the pixel where the maximum or minimum value of the eight neighbors is located; b) The number of pixels in the contour line Not less than 8; c) Vortex amplitude, that is, the difference between the SSHA value of the contour line and the value of the extreme point in the contour line, not less tha...

Embodiment 2

[0055] Anomalous mesoscale eddies were identified based on observational data of SSHA and SST remote sensing data in the global ocean (90°S-90°N, 180°W-180°E).

[0056] 1. Based on the HyperDense-Net model, the SSHA and SST data are cascaded in a densely connected manner to achieve the fusion of SST and SSHA features at different levels. HyperDense-Net is a network model developed based on densely connected networks to solve multimodal fusion ( figure 2 ), its forward propagation can be expressed by formula (1):

[0057]

[0058] In formula (1), x is a single network layer, the superscript s indicates which mode the network layer is in, and the subscript l indicates the network layer. function It is a compound operation function of batch normalization function (Batch Normalization, BN), activation function ReLU, and convolution operation (Conv). HyperDense-Net adds the direct connection of any layer to all subsequent layers in a forward manner, this dense connection no...

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 an ocean abnormal mesoscale eddy identification method based on deep learning and multi-source remote sensing data. The method comprises the following steps: firstly, constructing an abnormal mesoscale eddy sample library; performing feature extraction and learning on the SSHA data and the SSTA data by using a dense connection network, and fusing the SSHA data and the SSTAdata into fused feature data containing SSH and SST information; constructing an identification model by improving a U-Net image segmentation model; and training a recognition model by using the constructed abnormal mesoscale eddy sample library, and inputting the fused feature data into the recognition model for feature extraction, thereby realizing recognition of abnormal mesoscale eddy. According to the method, the fusion of the multi-source remote sensing data is realized by utilizing the multi-mode fusion network HyperDense-Net, the characteristics of the multi-source remote sensing dataare fully mined to realize information supplementation, and data support is provided for realizing more accurate and effective monitoring of the ocean mesoscale eddy anomaly.

Description

technical field [0001] The invention belongs to the technical field of ocean observation, and in particular relates to a method for identifying ocean anomaly mesoscale eddies based on deep learning. Background technique [0002] The mesoscale vortex is the intermediate link connecting the large scale and the small scale in the energy cascade. It is the link for transporting and mixing energy and matter in the ocean, and plays an important role in the distribution of ocean material energy and the energy exchange between the ocean and the atmosphere. There are tens of thousands of mesoscale eddies in the global ocean, which are divided into cyclonic eddies (rotating counterclockwise in the northern hemisphere) and anticyclonic vortices (rotating clockwise in the northern hemisphere). The cyclonic vortex (anticyclonic vortex) is usually accompanied by local divergence (convergence), resulting in upwelling (sinking current), making the sea surface temperature of the vortex lower...

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): G06T7/10G06T7/60G06K9/62G06N3/04G06N3/08
CPCG06T7/10G06T7/60G06N3/08G06T2207/10032G06T2207/20221G06T2207/20081G06T2207/20084G06T2207/30204G06N3/045G06F18/24
Inventor 刘颖洁李晓峰高乐任沂斌张旭东
Owner INST OF OCEANOLOGY - CHINESE ACAD OF SCI
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