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

Inductive radar high-resolution range profile recognition method based on deep transfer learning

A technology of high-resolution range profile and transfer learning, which is applied in the field of inductive radar high-resolution range profile recognition, can solve problems such as difficult application and real signal access, and achieve strong practicability, increased sample size, and strong portability Effect

Active Publication Date: 2018-12-11
XIAMEN UNIV
View PDF10 Cites 18 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Although the current mainstream radar high-resolution range image recognition method based on deep learning has achieved good results on simulation data, it is difficult to apply it in practice because of discrepancies in the response to real signals.

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
  • Inductive radar high-resolution range profile recognition method based on deep transfer learning
  • Inductive radar high-resolution range profile recognition method based on deep transfer learning
  • Inductive radar high-resolution range profile recognition method based on deep transfer learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0027] The following embodiments will further illustrate the present invention in conjunction with the accompanying drawings.

[0028] The present invention comprises the following steps:

[0029] 1) Preprocessing of the real target signal and auxiliary simulation data, the specific steps are as follows:

[0030] Step 1: In order to reduce the impact of intensity sensitivity and outliers on the algorithm, normalize the radar high-resolution range image signal data:

[0031]

[0032] Among them, x is the original one-dimensional radar high-resolution range image signal, ave(max 5 (x)) and ave(min 5 (x)) are the squares of the five maximum and minimum values ​​in the frame respectively, x * is the normalized signal data;

[0033] Step 2: In order to solve the problem that the dimensions of the simulated data are different from those of the real signal, firstly, the edge part of the real signal is counted, and according to its distribution characteristics, the noise points...

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 an inductive radar high-resolution range profile recognition method based on deep transfer learning, which relates to radar signal processing. The method includes preprocessing of a real target signal and aided simulation data; selection and optimization of a deep model; application of an inductive transfer strategy in deep transfer learning. Aiming at the difficulty of acquiring complete high-resolution range profile data, a target recognition framework based on deep transfer learning is proposed, which can effectively improve the recognition performance of radar high-resolution range profile with small sample size and incomplete attitude. Based on the transfer learning in deep learning and machine learning, the method is practical and portable, and can meet the needs of weak supervised learning in most small samples and incomplete situations.

Description

technical field [0001] The invention relates to radar signal processing, in particular to an inductive radar high-resolution range image recognition method based on deep transfer learning. Background technique [0002] As one of the core technologies to determine whether the weapon system is intelligent, automatic radar identification technology plays an important role in modern warfare. As an effective data source for target recognition, high-resolution one-dimensional range profile is easy to obtain and store, and is widely used in automatic radar target recognition. However, due to the attitude sensitivity, translation sensitivity, and intensity sensitivity of the signal itself, feature extraction has become a key technology in radar high-resolution range image target recognition. Deep learning uses end-to-end multi-layer nonlinear mapping to adaptively learn robust and identifiable features of the target through a large amount of data, and surpasses the performance of t...

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): G06K9/00G06N3/04G06N3/08
CPCG06N3/08G06N3/045G06F2218/04
Inventor 黄悦丁兴号余宪王继天文艺
Owner XIAMEN 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