An Adaptive Feature Learning Model for Sequential Radar High Resolution Range Profile Recognition

Peng, Xuan; Gao, Xueshan; Zhang, Yifan; Li, Xiang

doi:10.3390/s17071675

Cited by 16 publications

(13 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…All three classes of targets cover 0 to 360 degrees of aspect angles and their distance and azimuth resolutions are 0.3 m [ 43 , 44 ]. In the dataset, each target is obtained under the depression angle of 15° and 17°.…”

Section: Methodsmentioning

confidence: 99%

Attention-Based Recurrent Temporal Restricted Boltzmann Machine for Radar High Resolution Range Profile Sequence Recognition

Zhang

Gao

Peng

et al. 2018

Sensors

Self Cite

View full text Add to dashboard Cite

The High Resolution Range Profile (HRRP) recognition has attracted great concern in the field of Radar Automatic Target Recognition (RATR). However, traditional HRRP recognition methods failed to model high dimensional sequential data efficiently and have a poor anti-noise ability. To deal with these problems, a novel stochastic neural network model named Attention-based Recurrent Temporal Restricted Boltzmann Machine (ARTRBM) is proposed in this paper. RTRBM is utilized to extract discriminative features and the attention mechanism is adopted to select major features. RTRBM is efficient to model high dimensional HRRP sequences because it can extract the information of temporal and spatial correlation between adjacent HRRPs. The attention mechanism is used in sequential data recognition tasks including machine translation and relation classification, which makes the model pay more attention to the major features of recognition. Therefore, the combination of RTRBM and the attention mechanism makes our model effective for extracting more internal related features and choose the important parts of the extracted features. Additionally, the model performs well with the noise corrupted HRRP data. Experimental results on the Moving and Stationary Target Acquisition and Recognition (MSTAR) dataset show that our proposed model outperforms other traditional methods, which indicates that ARTRBM extracts, selects, and utilizes the correlation information between adjacent HRRPs effectively and is suitable for high dimensional data or noise corrupted data.

show abstract

Section: Methodsmentioning

confidence: 99%

Attention-Based Recurrent Temporal Restricted Boltzmann Machine for Radar High Resolution Range Profile Sequence Recognition

Zhang

Gao

Peng

et al. 2018

Sensors

Self Cite

View full text Add to dashboard Cite

show abstract

“…Compared with the twodimensional (2-D) synthetic aperture radar (SAR) images [1][2][3], on the one hand, HRRPs are easier to obtain, store, and process; on the other hand, HRRPs can maintain a certain stability when the target moves, while the SAR image tends to be blurred [4][5]. Therefore, HRRP-based target recognition has always been a hot topic in the field of radar automatic target recognition (ATR) [6][7][8]. For better recognition performance, it is a crucial issue that how to adequately learn the features from the HRRP data.…”

Section: Introductionmentioning

confidence: 99%

A Novel HRRP Target Recognition Method Based on LSTM and HMM Decision-making

Huang

Liu

et al. 2019

2019 25th International Conference on Automation and Computing (ICAC)

View full text Add to dashboard Cite

Feature learning is a key step of target recognition for high-resolution range profile (HRRP). In traditional methods, Hidden Markov Model (HMM) can learn the features of HRRP target aspect information. However, the contextual correlations of HRRPs are ignored due to the independence assumptions in HMM, which brings many limitations to feature learning and weakens the generalization performance. On the contrary, the Long Short-Term Memory (LSTM) network can learn the contextual correlations of HRRPs, but not the target aspect information. To overcome the limitations in these feature learning methods, a new HRRP target recognition method that combines LSTM with HMM decision-making is proposed in this paper. The method consists of two branches: one is target recognition based on HMMs directly; the other is that the latent correlations of HRRPs are extracted by LSTM network and then use HMMs to do target recognition. Finally, the recognition result is obtained by making joint decisions between the two branches. The HRRPs are generated by the inversion of real radar images and the experimental results show that the proposed algorithm outperforms both the HMM and LSTM method.

show abstract

“…But a large amount of training data is required for possible aspect angles. Recently, in [ 32 ], restricted Boltzmann machine was researched for sequential HRRP RATR, which is a supervised learning process and the training process is complicated with many hyper-parameters to tune. In this paper, we introduce the subspace decomposition to improve the robustness to the aspect sensitivity instead.…”

Section: Introductionmentioning

confidence: 99%

Robust Automatic Target Recognition via HRRP Sequence Based on Scatterer Matching

Jiang

Cai

et al. 2018

Sensors

View full text Add to dashboard Cite

High resolution range profile (HRRP) plays an important role in wideband radar automatic target recognition (ATR). In order to alleviate the sensitivity to clutter and target aspect, employing a sequence of HRRP is a promising approach to enhance the ATR performance. In this paper, a novel HRRP sequence-matching method based on singular value decomposition (SVD) is proposed. First, the HRRP sequence is decoupled into the angle space and the range space via SVD, which correspond to the span of the left and the right singular vectors, respectively. Second, atomic norm minimization (ANM) is utilized to estimate dominant scatterers in the range space and the Hausdorff distance is employed to measure the scatter similarity between the test and training data. Next, the angle space similarity between the test and training data is evaluated based on the left singular vector correlations. Finally, the range space matching result and the angle space correlation are fused with the singular values as weights. Simulation and outfield experimental results demonstrate that the proposed matching metric is a robust similarity measure for HRRP sequence recognition.

show abstract

An Adaptive Feature Learning Model for Sequential Radar High Resolution Range Profile Recognition

Cited by 16 publications

References 28 publications

Attention-Based Recurrent Temporal Restricted Boltzmann Machine for Radar High Resolution Range Profile Sequence Recognition

Attention-Based Recurrent Temporal Restricted Boltzmann Machine for Radar High Resolution Range Profile Sequence Recognition

A Novel HRRP Target Recognition Method Based on LSTM and HMM Decision-making

Robust Automatic Target Recognition via HRRP Sequence Based on Scatterer Matching

Contact Info

Product

Resources

About