Water body identification method based on deep dense neural network

A neural network and recognition method technology, applied in the field of water body recognition, can solve problems such as the inability to achieve efficient and accurate recognition, and achieve the effects of easy training and convergence, reducing gradient disappearance, and reducing the number of network parameters

Active Publication Date: 2019-07-26
CHENGDU UNION BIG DATA TECH CO LTD
View PDF5 Cites 15 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Efficient and accurate identification of water bodies, especially small rivers, cannot be achieved

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
  • Water body identification method based on deep dense neural network
  • Water body identification method based on deep dense neural network
  • Water body identification method based on deep dense neural network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0031] In order to make the purpose, technical solution and advantages of the present invention clearer, the present invention will be further elaborated below in conjunction with the accompanying drawings.

[0032] In this example, see figure 1 As shown, the present invention proposes a water body recognition method based on a deep dense neural network, comprising steps:

[0033] S100, data collection, downloading satellite remote sensing image data, and marking water body and non-water body parts in the image data;

[0034] S200, establishing a dense UNet segmentation network model;

[0035] S300, using the labeled training set data to optimize and train the dense UNet segmentation network model;

[0036] S400, input the test set data into the optimized network model, and identify the water body area in the test set image.

[0037] As the optimization scheme of the foregoing embodiment, the establishment of the dense UNet segmentation network model includes steps:

[003...

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 water body identification method based on a deep dense neural network, which comprises the following steps: data acquisition: downloading satellite remote sensing image data,and labeling water body and non-water body parts in the image data; establishing an intensive UNet segmentation network model; carrying out optimization training on the intensive UNet segmentation network model by using the labeled training set data; inputting the test set data into the optimized network model, identifying the water body area in the test set image, and verifying the model effect.According to the method, the parameters of the neural network for water body identification can be effectively reduced on the premise of ensuring the accuracy, the training time is greatly shortened,and the difficulty of a real-time environment monitoring task of remote sensing is greatly reduced.

Description

technical field [0001] The invention belongs to the technical field of water body recognition, in particular to a water body recognition method based on a deep dense neural network. Background technique [0002] At present, satellite remote sensing is widely used in various aspects such as environmental monitoring, weather forecasting, and disaster prevention, among which water body identification is an important application of satellite remote sensing. Water body identification is the premise of water body pollution detection, and its accuracy directly affects the subsequent calculation of pollutant content and pollution range. It is an important part of the application of satellite remote sensing and automatic environmental monitoring. [0003] Existing deep learning methods such as convolutional neural networks perform semantic segmentation to extract water body parts in remote sensing images. However, while the large-scale convolutional neural network increases the reco...

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): G06K9/00G06N3/04G06N3/08
CPCG06N3/084G06V20/13G06N3/045
Inventor 不公告发明人
Owner CHENGDU UNION BIG DATA TECH CO LTD
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