Unlock instant, AI-driven research and patent intelligence for your innovation.

Sea surface target detection method based on stack auto-encoder

A self-encoder and target detection technology, which is applied in the directions of instruments, character and pattern recognition, scene recognition, etc., can solve the problems of easy missing detection and high SNR requirements

Active Publication Date: 2020-05-12
NANJING UNIV OF AERONAUTICS & ASTRONAUTICS
View PDF2 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] Aiming at the problem that traditional CA-CFAR detection has high requirements on SNR and is prone to missed detection, the purpose of the present invention is to provide a sea surface target detection method based on a stacked autoencoder

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
  • Sea surface target detection method based on stack auto-encoder
  • Sea surface target detection method based on stack auto-encoder
  • Sea surface target detection method based on stack auto-encoder

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0068] Use two-dimensional K distribution clutter and targets with different SNRs to build a sea surface target model, simulate the echo data after pulse compression, use the VV polarization method, select the radar parameters and environmental parameters as shown in Table 1, and simulate a 500×500 second Dimensional K distribution sea clutter such as figure 2 Shown.

[0069] Table 1

[0070]

[0071] Randomly select the 150th distance unit, and its amplitude and spectrogram are as follows image 3 As shown, it obeys the K distribution model. Add 10 random-sized targets at random locations in the sea clutter. Here, it is assumed that the target azimuth does not exceed 2 units, and the distance does not exceed 25 units. The target signal to noise ratio (SCR) of each unit is: Where P s Is the target power of each unit, P c Is the background clutter power. Reserve a certain protection unit, select the average power of M reference units around the unit under test to estimate P c . ...

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 sea surface target detection method based on a stack auto-encoder, and the method comprises the following steps: 1, carrying out the preprocessing of sea surface echo data, obtaining a two-dimensional image, and dividing the two-dimensional image into a training set and a test set according to the azimuth time; 2, sea surface target detection is carried out based on a stack auto-encoder, the stack auto-encoder is formed by sequentially connecting an input layer, two hidden layers composed of auto-encoders and a logistic classifier, and sea surface target detection comprises the following three steps of layer-by-layer unsupervised pre-training, supervised fine adjustment and test of a test set; and 3, performing performance test on the network model trained in thestep 2 by using the test set. According to the invention, false alarm can be effectively suppressed, the detection rate is improved, and the method has better detection performance than traditional CA-CFAR detection under the condition of low SNR.

Description

Technical field [0001] The invention relates to a radar signal processing algorithm combined with deep learning, in particular to a sea surface target detection method based on a stacked autoencoder (SAE), belonging to the technical field of radar target detection. Background technique [0002] The detection and classification of sea targets are widely used in military and civilian fields. Radar is an important means of detection and surveillance of sea targets. However, it is reliable and robust due to the sea clutter generated by the complex marine environment and the diversified types of sea targets. Sea surface target detection and classification is always one of the key technologies that need to be studied. The focus of the current research on the statistical characteristics of sea clutter is the parameter estimation and simulation of the composite K distribution. For details, see the literature "Zhao Haiyun, Hu Xuecheng. Spatial-temporal two-dimensional correlation K distri...

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): G06K9/00G06K9/62
CPCG06V20/13G06F18/24G06F18/214Y02A90/10
Inventor 闫贺王珏黄佳陈超李瑞安
Owner NANJING UNIV OF AERONAUTICS & ASTRONAUTICS