SAR original data target recognition method based on ResNet18
A target recognition and original data technology, applied in character and pattern recognition, instruments, biological neural network models, etc., can solve the problems of old CNN architecture, small scale, and affecting the accuracy of target recognition, and achieve low complexity and network training Easier to optimize and identify better classification performance
Active Publication Date: 2022-02-15
NANJING UNIV OF AERONAUTICS & ASTRONAUTICS
View PDF14 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
The first framework works directly in the phase history domain, while the other framework includes three steps of image reconstruction, dephasing the image and returning it to the phase history domain, but the input part of the method utilizes an approximation of the phase history instead of The actual value often affects the accuracy of target recognition
In addition, the CNN architecture used is relatively old and small
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 moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment
[0065] In order to verify the effectiveness of ResNet18-based SAR raw data target recognition, experimental verification is carried out on the simulated data set and the MSTAR data set. In addition, in order to illustrate the advantages of ResNet18-based SAR raw data target recognition, the target recognition results obtained by this method are compared with ResNet18-based SAR image target recognition results and SAR raw data target recognition results based on other CNNs.
[0066] Figure 6 In order to simulate the network training process of the data set, the network training time is 4 minutes and 26 seconds, and the recognition accuracy rate is 99.6%. It can be seen from the results that under the condition of simple and small samples, the network using SAR received signals as input The model recognition rate reaches over 99%, the network training efficiency is high and the target classification can be achieved well.
[0067] Figure 7 It is the network training process o...
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
Login to View More Abstract
The invention discloses an SAR original data target recognition method based on an ResNet18, relates to the technical field of radar signal processing, and provides an automatic target recognition (ATR) method with synthetic aperture radar (SAR) original data as input. According to an existing ATR method, a task is executed after an image is formed. However, in an imaging process of target information, target abstract feature information hidden in original data may be lost, so that the recognition accuracy is limited. Therefore, according to the invention, on the basis of a deep residual network ResNet18, the SAR original data is sent to a convolutional neural network framework which does not need image reconstruction for target recognition and classification, the recognition efficiency is obviously improved, and a better classification result is obtained.
Description
technical field [0001] The invention relates to the technical field of radar signal processing, in particular to a ResNet18-based SAR raw data target recognition method. Background technique [0002] With the rapid development of modern information technology and its wide application in the military field, target recognition technology has a wide range of applications in military fields such as early warning detection, precision guidance, battlefield command and identification. Automatic Target Recognition (ATR) of Synthetic Aperture Radar (SAR) refers to the task of finding and identifying targets, and its purpose is to identify target features by acquiring attribute information in images. However, the existing SAR target recognition technology generally has the disadvantages of low intelligence and poor real-time performance. Therefore, research on smarter and more efficient target recognition technology has become an important demand. [0003] In recent years, due to th...
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
Login to View More IPC IPC(8): G06V20/10G06V10/774G06V10/82G06K9/62G06N3/04
CPCG06N3/045G06F18/214
Inventor 汪玲阮西玥郭军胡长雨
Owner NANJING UNIV OF AERONAUTICS & ASTRONAUTICS