Small target detection method on sea surface based on deep learning of time-frequency map

A technology of small target detection and deep learning, which is applied in the field of shore-based or ship-borne high-resolution radar to comprehensively improve the performance of sea detection, can solve the problems of unmeasured data verification and algorithm false alarm control, and achieve the effect of improving the detection probability

Active Publication Date: 2022-07-05
NANJING UNIV OF INFORMATION SCI & TECH
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

In addition, the convolutional neural network method was introduced into the detection and classification of sea surfac

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
  • Small target detection method on sea surface based on deep learning of time-frequency map
  • Small target detection method on sea surface based on deep learning of time-frequency map
  • Small target detection method on sea surface based on deep learning of time-frequency map

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0047] refer to figure 1 The invention solves the problem of artificial feature extraction limitation and false alarm control, and realizes robust detection of small sea surface targets in different detection environments. Specific steps are as follows:

[0048] Step 1, get the echo vector z m

[0049] 1.1) For shore-based radar or shipborne radar in dwell mode, the radar receives a two-dimensional data matrix of N pulses of M distance units, denoted as R(n,m), n=1,2,... ,N; m=1,2,...,M; where, n represents the pulse dimension, and m represents the distance dimension; assuming that the target does not have a distance extension, each distance unit is independently detected. Suppose the echo vector of any one of the distance units is z m =[R(1,m),R(2,m),...,R(N,m)] T , called the cell undertest (CUT). In the distance dimension, take the echo vectors of K reference distance units around the CUT, denoted as z m,k ,k=1,2,...,K,K

[0050] 1.2) The essence of detection is ...

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 present invention provides a detection method based on time-frequency map deep learning, aiming at the problem that the echo samples containing the target are sparse and incomplete, and the artificially extracted features are often empirical and qualitative. The invention obtains a large amount of data including target echoes by a semi-simulation method, and solves two types of data imbalance problems. Through the whitening preprocessing and transfer learning classifier, the characteristics of the time-frequency map can be learned autonomously, and the difference between the two types of data can be deeply mined. The invention uses the two types of probabilities output by the classifier as statistics, and solves the problem of false alarm control, and this constant false alarm characteristic has important significance in actual radar detection.

Description

Technical field: [0001] The design of the invention belongs to the technical field of radar signal processing, and relates to a detection method for small targets on the sea surface under low signal-to-noise ratio, which can be used for shore-based or shipborne high-resolution radars to comprehensively improve the sea detection performance. Background technique: [0002] With the development of stealth technology and target miniaturization, the main task of high-resolution sea radar is to detect small targets on the sea surface in complex environments, such as boats, speedboats, submarines, and floating objects. Usually, the echoes of these small targets are submerged in the background of strong time-varying sea clutter and have a low signal-to-noise ratio. Compared with air targets, low-speed sea surface targets usually have weaker maneuverability, which leads to their frequency spectrum easily falling in the main clutter region of sea clutter, which increases the difficult...

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): G01S13/937G01S7/41G01S7/40G06K9/62G06N3/04G06N3/08
CPCG01S13/937G01S7/414G01S7/4052G06F18/24
Inventor 施赛楠董泽远杨静
Owner NANJING UNIV OF INFORMATION SCI & TECH
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