Multiview Synthetic Aperture Radar Automatic Target Recognition Optimization: Modeling and Implementation

Pei, Jifang; Huang, Yulin; Sun, Zhichao; Zhang, Yin; Yang, Jianyu; Yeo, Tat‐Soon

doi:10.1109/tgrs.2018.2838593

Cited by 27 publications

(19 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…However, based on the analysis above, traditional convex optimization methods are not applicable to the multiple objectives problem. To solve the CMOP, the nonlinear constrained nondominated sorting genetic algorithm II (NC-NSGA II) is adopted in this paper [47], [48]. NC-NSGA II is a fast and effective intelligent evolutionary algorithm.…”

Section: Gc Optimization Design Solutionmentioning

confidence: 99%

Forward-Looking Geometric Configuration Optimization Design for Spaceborne-Airborne Multistatic Synthetic Aperture Radar

Mao

Pei

Huo

et al. 2021

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

Self Cite

View full text Add to dashboard Cite

Spaceborne-airborne multistatic synthetic aperture radar (SA-MuSAR) has the ability to provide high-resolution forward-looking imagery for receivers, but it relies on careful design of the geometric configuration (GC). In this paper, a forward-looking GC optimization design method is proposed to obtain a high-quality fused image with limited observation time. First, the relationship between the spatial resolution and GC is illustrated by the wavenumber spectrum distribution of SA-MuSAR. Second, GC evaluators depending on the distribution of multiple wavenumber spectrum data are proposed. The GC design problem of coherent SA-MuSAR is transformed into a constrained multi-objective optimization problem (CMOP). An intelligent evolutionary algorithm is adopted to optimize the wavenumber spectrum distribution. With the proposed method, high-quality forward-looking imagery can be obtained with a short observation time. Numerical simulations are carried out to verify the effectiveness of the proposed method.

show abstract

Section: Gc Optimization Design Solutionmentioning

confidence: 99%

Forward-Looking Geometric Configuration Optimization Design for Spaceborne-Airborne Multistatic Synthetic Aperture Radar

Mao

Pei

Huo

et al. 2021

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

Self Cite

View full text Add to dashboard Cite

show abstract

“…由于仿真样本与真实样本处于不同的数据域中, 直接将仿真样本作为扩充样本未必能够提升分类器的性能, Liu 等 [39] 指出在 B52 和 B707 两类 SAR 飞机识别问题中仿真样本的加入不但没有提升分类器的性能反而使得识别率有所下降, 并将原因归结为 SAR 真实样本与仿真样本在图像域的差异. 针对这一问题, Liu 等采用循环生成对抗网络 CycleGAN [40] 利用多角度 SAR 图像进行目标识别 [51,52] 是近年来的一项研究热点, Zhang 等 [53] 通过双向长短期记忆 (long short-term memory, LSTM) 循环神经网络利用多角度图像对 SAR 目标进行分类, 方法的原理图请参考文献 [53] 中的图 1, 首先选取具有不同方位角的目标图像组成多角度目标图像序列, 然后采用 Gabor 滤波器提取特征并利用局部二值模式 (local binary pattern, LBP) 对特征进行编码, 接下来用一个浅层的神经网络对特征进行降维, 最后利用双向循环神经网络将多角度信息进行融合从而计算目标的类别. 实验结果表明, 方法在 MSTAR 数据集中的 10 分类问题上能够达到 99% 的识别率, 且具有更好的抗噪和抗混淆性能.…”

Section: 近年来随着深度学习理论的不断发展相关方法已在众多领域取得突破性进展卷积神经网络unclassified

Progress of deep learning-based target recognition in radar images

Pan

Zhang

2019

Sci. Sin.-Inf.

View full text Add to dashboard Cite

Deep-learning-based extraction of the animal migration patterns from weather radar images SCIENCE CHINA Information Sciences 63, 140304 (2020); Deep learning approach for recognition and classification of yield affecting paddy crop stresses using field images Aritificial Intelligence in Agriculture 4, 12 (2020); Radar automatic target recognition based on feature extraction for complex HRRP Science in China Series F-Information Sciences 51, 1138 (2008); Deep reinforcement learning-based joint task offloading and bandwidth allocation for multiuser mobile edge computing 数字通信和网络 5, 10 (2019); Multitask deep learning-based multiuser hybrid beamforming for mm-wave orthogonal frequency division multiple access systems

show abstract

“…The experiment was carried out using MATLAB software [15] on a laboratory server with configuration of Windows 7 system, I7 processor and 16G memory.…”

Section: Experimental Environmentmentioning

confidence: 99%

Airport Target Recognition and Ground Detection Management Based on Neural Network

2019

IJOMAM

View full text Add to dashboard Cite

Modern airport management system is complex. For airport managers, it is increasingly difficult to identify aircraft and airport ground. High-resolution high-altitude reconnaissance equipment provides an effective tool for airport supervision. In order to effectively recognize aircraft and airport ground in reconnaissance image, this paper briefly introduced convolutional neural network (CNN) and airport target recognition model based on CNN. In order to improve the recognition and location accuracy of the recognition model, the loss function of the model was improved. Then, two recognition models were simulated and analyzed by using MATLAB software. It was found that the improved recognition model was better than the former one in the aspect of visual effect of recognition; the improved model was closer to the real value in traffic density recognition of airport operation state; the accuracy and recall rate of dispatch management opinions based on traffic density feedback were higher; the time required for recognition and management was shorter. In summary, the improved identification management model based on CNN can effectively identify airport aircraft, and quickly feedback accurate management dispatching opinions according to the identified airport operation status.

show abstract

Multiview Synthetic Aperture Radar Automatic Target Recognition Optimization: Modeling and Implementation

Cited by 27 publications

References 31 publications

Forward-Looking Geometric Configuration Optimization Design for Spaceborne-Airborne Multistatic Synthetic Aperture Radar

Forward-Looking Geometric Configuration Optimization Design for Spaceborne-Airborne Multistatic Synthetic Aperture Radar

Progress of deep learning-based target recognition in radar images

Airport Target Recognition and Ground Detection Management Based on Neural Network

Contact Info

Product

Resources

About