Water surface and underwater target classification method based on interference fringes and deep learning

A technology of interference fringes and deep learning, applied in the fields of marine engineering and underwater acoustic engineering

Active Publication Date: 2019-06-25
NORTHWESTERN POLYTECHNICAL UNIV
View PDF5 Cites 18 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Under the condition that the sound speed profile SSP (sound speed profile) is known, a large number of sound field interference fringe images can be obtained thr

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 surface and underwater target classification method based on interference fringes and deep learning
  • Water surface and underwater target classification method based on interference fringes and deep learning
  • Water surface and underwater target classification method based on interference fringes and deep learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0044] Now in conjunction with embodiment, accompanying drawing, the present invention will be further described:

[0045] The invention proposes to use the Gaussian ray acoustic model BELLHOP and the known SSP simulation to obtain thousands of sound field interference fringe images with different sound source depths, including complete images without interference, fuzzy images and images with random depth errors. Then the blurred image and the image with random depth error are used as the input training set, and the complete image without interference is used as the output training set to train DBN; all the stripe images are used as the input training set, and the category of the image is used as the output training set to train CNN, DBN and CNN needs to be trained thousands of times until it converges. After the training of DBN and CNN, DBN is used as the front-end processing module of CNN, and beamforming is added as a preprocessing module. Finally, a judgment module is add...

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 relates to a water surface and underwater target classification method based on interference fringes and deep learning. Firstly, a Gaussian beam acoustic model is used for simulation toobtain a large number of acoustic field interference fringe images, which comprise a clear image, a blurred image and an image with random deep errors, as a training set for a deep belief network (DBN) and a CNN; the well-trained DBN is then used as a front-end processing module of the CNN, and beams are added at the same time to form a pre-processing module of the system; and finally, a pre-processed image is inputted to the DBN, and the inputted fringe image is thus autonomously classified. The method solves the problem that the training set sample is too small in deep learning by simulatinga large number of acoustic field interference fringe images. The DBN model is designed to optimize the blur and fringe shift problems in the actual input image. The CNN is designed to autonomously classify water surface and underwater targets.

Description

technical field [0001] The invention belongs to the fields of underwater acoustic engineering, marine engineering technology, etc., and relates to a method for classifying objects on the surface and underwater based on interference fringes and deep learning. A classification method based on interference fringe images and deep neural networks DBN and CNN to classify surface and underwater targets, suitable for surface targets with a target depth of more than 20m and underwater targets below 20m. Background technique [0002] In the field of image recognition, the classification of surface and underwater objects has always been an urgent problem to be solved. The traditional method is to use signal processing methods such as beamforming and DOA (beamforming and direction of arrival) estimation to estimate the source depth, and then classify the surface and underwater targets according to the estimated target depth. Although this method can describe the physical characteristic...

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): G01S7/539
Inventor 杨坤德周星月
Owner NORTHWESTERN POLYTECHNICAL UNIV
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