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

Deep learning-based radar echo signal target identification method

A radar echo and target recognition technology, applied in the field of target signal recognition, can solve the problems of imaging preprocessing and low recognition accuracy, and achieve the effect of improving recognition accuracy, high recognition accuracy, and fast training speed

Active Publication Date: 2018-06-29
SOUTHEAST UNIV
View PDF7 Cites 54 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0007] Technical solution: In order to solve the problem that the existing SAR image-based automatic target recognition system needs imaging preprocessing and the recognition accuracy is not high, the present invention proposes a radar echo signal automatic target recognition method based on deep learning, including the following steps:

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-based radar echo signal target identification method
  • Deep learning-based radar echo signal target identification method
  • Deep learning-based radar echo signal target identification method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0049] Considering the recognition of six types of simple targets, the six types of targets used for recognition are cubes, triangular prisms, tetrahedrons, spheres, cylinders and cones, such as figure 2 shown. The echo amplitudes of six types of targets and sample images of SAR images are also figure 2 are given separately.

[0050] Using the radar echo signal classification method based on deep learning in the present invention to classify the radar echo signals to be classified of the test set includes the following steps:

[0051] (1) Perform (0,1) normalization operation on the radar echo signal to be classified, and select the 220×220 sub-block in the data center of the original echo signal as the sample to be classified;

[0052] (2) Use balanced sample training, that is, the number of training samples in each category is equal, and each of the six categories takes 400 as training samples. The technique of increasing the amount of data is adopted for the training s...

Embodiment 2

[0069] Considering the identification of six types of missile targets, the six types of missile targets used for identification are as follows Figure 4 Classes (a)-(f) shown. The echo amplitudes of six types of missile targets, SAR images and SAR original imaging data are also included in the Figure 4 (g)-(l), (m)-(r) and (s)-(x), respectively. Use the trained convolutional neural network to test on the test set. The recognition results are shown in Table 3. The classification results in Table 3 show that the target recognition accuracy based on echo data reaches 100%, and the target recognition based on SAR images is accurate. The rate is 98.83%, while the recognition accuracy based on SAR original imaging data is 99.83%. It can be seen from Table 3 that when training based on SAR images and SAR original imaging data, the recognition accuracy of Missile 3 is low, 93% and 99% respectively. In order to better illustrate the performance difference between the echo data set,...

Embodiment 3

[0081]The SAR imaging algorithm realizes the high resolution of SAR through pulse compression in the range direction and matched filtering in the azimuth direction. Therefore, there is a large loss of information during the imaging process. In order to prove that the radar echo signal contains more information than the SAR image, the information entropy of the echo signal and the SAR image is compared. Consider the SAR images and radar echo signals of cubes and missile targets respectively, such as Figure 6 As shown, it can be seen that the echo data contains richer information than the corresponding SAR image, which indicates that the echo data has greater information entropy. It is calculated that the information entropy of cube target echo data and SAR image is 4.24 and 0.22 respectively; the information entropy of missile model echo data and SAR image is 5.61 and 0.32 respectively. This result is in line with the intuitive feeling, which proves that the echo data used i...

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 deep learning-based radar echo signal target identification method. The method directly identifies radar echo signals, consequently, information loss in the process of complex two-dimensional matched filtering processing and imaging of an SAR (synthetic aperture radar) image is prevented, and the accuracy of identification is effectively improved. In addition, the methodutilizes a convolutional neural network, thus avoiding the complex preprocessing and feature extraction process for echo signals and greatly simplifying the identification processing process. The deeplearning-based radar echo signal target identification method mainly solves the problem that imaging preprocessing is required by a conventional SAR image-based target identification method; applyingthe convolutional neural network method in original radar echo data, the deep learning-based radar echo signal target identification method has the advantages of high identification accuracy and goodanti-noise performance, and moreover, the deep learning-based radar echo signal target identification method has higher effectiveness and practicability.

Description

technical field [0001] The invention belongs to target signal recognition, in particular to a radar echo signal target recognition method based on deep learning. Background technique [0002] Remote sensing technology is a comprehensive technology that uses various sensing instruments to collect and process electromagnetic wave information radiated and reflected by long-distance targets based on electromagnetic wave theory, so as to detect and identify various scenes on the ground and sea. Remote sensing technology collects electromagnetic radiation information of ground objects from artificial satellites, aircraft or other aircraft. These information contain a large amount of metadata, so that we can easily retrieve geographic parameters, biochemical quantities, and detect and classify targets. Such detection and classification requires powerful statistical methods, in which deep learning algorithms play a key role. [0003] Synthetic aperture radar is an active earth obse...

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): G06K9/00G06K9/62G06N3/04G06N3/08
CPCG06N3/08G06V20/13G06N3/045G06F18/24147
Inventor 崔铁军范湉湉
Owner SOUTHEAST 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