Cadence Velocity Diagram Research Articles

Sample Intercorrelation-Based Multidomain Fusion Network for Aquatic Human Activity Recognition Using Millimeter-Wave Radar

To address the issue that existing multi-domain fusion methods do not consider data correlation within mini-batch data, the first attempt is made in this letter to propose a method based on sample inter-correlation learning multi-domain fusion network (SIMFNet), which aims to accommodate multivariate domain data and further enhance the aquatic human activity recognition performance. To fully utilize the radar multidimensional information, the three-branch convolution neural network (CNN) feature extractor is first employed to extract domain-specific features from the time-range map (TRM), time-Doppler map (TDM) and cadence velocity diagram (CVD). Then, the multi-domain features are fused and fed into the graph construction layer (GCL) to generate instance graphs. Next, a graph aggregation layer (GAL) is applied to aggregate node information from various-hop neighborhood domains. Finally, node-level classification is used to achieve aquatic human activity recognition. The experimental results evaluated on the built aquatic human activity recognition dataset demonstrate that the proposed SIMFNet has better generalization performance than the state-of-the-art multi-domain fusion methods.

Wheeled Vehicle Detection Under the Asymmetric Micro-Doppler Interference

With the development of automated driving (AD) applications, the micro-Doppler effect induced by the rolling wheels of the vehicles normally disturbs the target detection and the effect is difficult to be eliminated at close range. In this article, by illustrating the micro-Doppler motion of the rolling wheels, the asymmetric Doppler spectrum caused by the micro-Doppler interference is innovatively introduced. A random-finite-set (RFS)-based measurement model is applied to analyze the asymmetry. To eliminate the asymmetric phenomenon, a temporal smoothing method based on the cadence velocity diagram (CVD) is proposed. Simulations based on the RFS model and experiments prove the exhibition of the asymmetric micro-Doppler interference introduced by rolling wheels and the feasibility of the proposed elimination method. Moreover, compared with the state-of-the-art symmetry method, the detection probability improves from 80% to 93% when the false alarm rate is fixed at 10 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">−6</sup> , showing that the proposed method is highly efficient and the false detection caused by the micro-Doppler effect can be eliminated accurately. The results may provide a new way for high-quality wheeled vehicle detection.

RCS Estimation of Drone Motion Using Mesh-Element Rotation in MoM and Micro-Doppler Signatures

For an automatic recognition system, a huge number of micro-Doppler signature images is required to train a classifier. However, in general, measurements are only possible for hovering drones, so the dynamic radar cross section (RCS) of a drone with various motions is estimated through EM simulation and converted into a micro-Doppler signature image. In this work, we use far-field approximation and mesh-element rotation in the method of moments to quickly estimate the dynamic RCS of a drone in various motions. First, 3-D meshes and the impedance matrix of a single propeller are generated only once. Instead of rotating the propeller, the position information of the mesh elements is rotated without recreating 3-D meshes or the impedance matrix. Next, using the mirror-image symmetric characteristics of drone propellers and the far-field approximation, the dynamic RCS of multiple propellers is synthesized from that of a single propeller. Finally, the dynamic RCS was estimated at 3 and 9 GHz and converted to micro-Doppler signatures, spectrogram, and cadence-velocity diagram (CVD). As a result, it was more accurate to estimate the dynamic RCS at 9 GHz, and it was easier to obtain the rotation frequency at 3 GHz in the CVD.

Space Target Classification Improvement by Generating Micro-Doppler Signatures Considering Incident Angle.

Classifying space targets from debris is critical for radar resource management as well as rapid response during the mid-course phase of space target flight. Due to advances in deep learning techniques, various approaches have been studied to classify space targets by using micro-Doppler signatures. Previous studies have only used micro-Doppler signatures such as spectrogram and cadence velocity diagram (CVD), but in this paper, we propose a method to generate micro-Doppler signatures taking into account the relative incident angle that a radar can obtain during the target tracking process. The AlexNet and ResNet-18 networks, which are representative convolutional neural network architectures, are transfer-learned using two types of datasets constructed using the proposed and conventional signatures to classify six classes of space targets and a debris–cone, rounded cone, cone with empennages, cylinder, curved plate, and square plate. Among the proposed signatures, the spectrogram had lower classification accuracy than the conventional spectrogram, but the classification accuracy increased from 88.97% to 92.11% for CVD. Furthermore, when recalculated not with six classes but simply with only two classes of precessing space targets and tumbling debris, the proposed spectrogram and CVD show the classification accuracy of over 99.82% for both AlexNet and ResNet-18. Specially, for two classes, CVD provided results with higher accuracy than the spectrogram.

Classification of Space Objects by Using Deep Learning with Micro-Doppler Signature Images.

Radar target classification is an important task in the missile defense system. State-of-the-art studies using micro-doppler frequency have been conducted to classify the space object targets. However, existing studies rely highly on feature extraction methods. Therefore, the generalization performance of the classifier is limited and there is room for improvement. Recently, to improve the classification performance, the popular approaches are to build a convolutional neural network (CNN) architecture with the help of transfer learning and use the generative adversarial network (GAN) to increase the training datasets. However, these methods still have drawbacks. First, they use only one feature to train the network. Therefore, the existing methods cannot guarantee that the classifier learns more robust target characteristics. Second, it is difficult to obtain large amounts of data that accurately mimic real-world target features by performing data augmentation via GAN instead of simulation. To mitigate the above problem, we propose a transfer learning-based parallel network with the spectrogram and the cadence velocity diagram (CVD) as the inputs. In addition, we obtain an EM simulation-based dataset. The radar-received signal is simulated according to a variety of dynamics using the concept of shooting and bouncing rays with relative aspect angles rather than the scattering center reconstruction method. Our proposed model is evaluated on our generated dataset. The proposed method achieved about 0.01 to 0.39% higher accuracy than the pre-trained networks with a single input feature.

Detection of multiple micro‐drones via cadence velocity diagram analysis

Many studies have demonstrated the capability of radar micro-Doppler signature for classifying micro-drones. However, most existing works on radar classification of drones are based on the assumption that the received signal is only reflected from a single drone. When multiple drones are present simultaneously, the existing methods of drone classification fail due to the superimposition of the micro-Doppler features of multiple drones. In this Letter, a method for detection of multiple drones is proposed. First the time–frequency spectrogram is converted into the cadence-velocity diagram (CVD), which expresses how the curves in the time-frequency-domain repeat. Then the cadence frequency spectrum (CFS), as the basis vector of the training data from each class, of the CVD is extracted. Finally, the K-means classifier is used to recognise the component of multiple micro-drones based on the CFS. The experimental results on real radar data demonstrate that the proposed method is capable of dealing with multiple drones with satisfactory classification accuracy.

Regularized 2-D complex-log spectral analysis and subspace reliability analysis of micro-Doppler signature for UAV detection

Unmanned aerial vehicle (UAV) has become an important radar target recently because of its wide applications and potential security threats. Traditionally, visual features such as spectrogram were often extracted for human operators to identify the micro-Doppler signature (mDS) of UAVs, i.e. sinusoidal modulation. In this paper, the authors aim to design a system for machine automatic classification of UAVs from other targets, particularly from birds as both UAVs and birds are small and slow-moving radar targets. Most existing mDS representations such as spectrogram, cepstrogram and cadence velocity diagram discard the phase spectrum, and only make use of the magnitude spectrum. What’s more, people often take the logarithm of the spectrum to enlarge the weak mDS but without sufficient care, as noise may be enlarged at the same time. The authors thus propose a regularized 2-D complex-log-Fourier transform to address these problems. Furthermore, the authors propose an object-oriented dimension-reduction technique: subspace reliability analysis, which directly removes the unreliable feature dimensions of two class-conditional covariance matrices in two separate subspaces. On the benchmark dataset, the proposed approach demonstrates better performance than the state-of-the-art approaches. More specifically, the proposed approach significantly reduces the equal error rate of the second best approach, cadence velocity diagram, from 6.68% to 3.27%.

Drone Classification Using Convolutional Neural Networks With Merged Doppler Images

We propose a drone classification method based on convolutional neural network (CNN) and micro-Doppler signature (MDS). The MDS only presents Doppler information in time domain. The frequency domain representation of MDS is called as cadence-velocity diagram (CVD). To analyze the Doppler information of drone in time and frequency domain, we propose a new image by merging MDS and CVD, as merged Doppler image. GoogLeNet, a CNN structure, is utilized for the proposed image data set because of its high performance and optimized computing resources. The image data set is generated by the returned Ku-band frequency modulation continuous wave radar signal. Proposed approach is tested and verified in two different environments, anechoic chamber and outdoor. First, we tested our approach with different numbers of operating motor and aspect angle of a drone. The proposed method improved the accuracy from 89.3% to 94.7%. Second, two types of drone at the 50 and 100 m height are classified and showed 100% accuracy due to distinct difference in the result images.

Recognition of humans based on radar micro‐Doppler shape spectrum features

In this study, a feature extraction algorithm is presented which automatically generates a set of shape spectrum features based on the cadence velocity diagram of the human micro-Doppler signature. Recognition performance between humans undertaking the same activity is assessed on a set of experimental data collected with a continuous wave radar operating at X-band using a Naive Bayesian classifier and a shape-similarity-spectrum classifier. Recognition performance is analysed as a function of key parameters, such as the dwell time on the target and the size of the training set, to investigate the level of robustness of the proposed features. Results show that high level recognition performance can be achieved for both the walking and running activities.

A novel algorithm for radar classification based on doppler characteristics exploiting orthogonal Pseudo-Zernike polynomials

Phase modulation induced by target micromotions introduces sidebands in the radar spectral signature returns. Time-frequency distributions facilitate the representation of such modulations in a micro-Doppler signature that is useful in the characterization and classification of targets. Reliable micro-Doppler signature classification requires the use of robust features that are capable of uniquely describing the micromotion. Moreover, future applications of micro-Doppler classification will require meaningful representation of the observed target by using a limited set of values. In this paper, the application of the pseudo-Zernike moments for micro-Doppler classification is introduced. Specifically, the proposed algorithm consists of the extraction of the pseudo-Zernike moments from the cadence velocity diagram (CVD). The use of pseudo-Zernike moments allows invariant features to be obtained that are able to discriminate the content of two-dimensional matrices with a small number of coefficients. The analysis has been conducted both on simulated and on real radar data, demonstrating the effectiveness of the proposed approach for classification purposes.