Z. Bao

SRMF-CLEAN Imaging Algorithm for Space Debris

A single-range matching filtering (SRMF)-CLEAN imaging algorithm for space debris is proposed in this paper. The proposed algorithm can successfully incorporate the prior moving information of the space debris into the imaging process. Since the SRMF algorithm can improve the resolution and imaging speed effectively, and the disturbance of the point spread function can be cancelled by the modified CLEAN algorithm, the high resolution images of space debris can be obtained by SRMF-CLEAN method effectively. Studies and simulations confirm the validness of the proposed algorithm.

A kernel optimization method based on the localized kernel Fisher criterion

It is widely recognized that whether the selected kernel matches the data determines the performance of kernel-based methods. Ideally it is expected that the data is linearly separable in the kernel induced feature space, therefore, Fisher linear discriminant criterion can be used as a cost function to optimize the kernel function. However, the data may not be linearly separable even after kernel transformation in many applications, e.g., the data may exist as multimodally distributed structure, in this case, a nonlinear classifier is preferred, and obviously Fisher criterion is not a suitable choice as kernel optimization rule. Motivated by this issue, we propose a localized kernel Fisher criterion, instead of traditional Fisher criterion, as the kernel optimization rule to increase the local margins between embedded classes in kernel induced feature space. Experimental results based on some benchmark data and measured radar high-resolution range profile (HRRP) data show that the classification performance can be improved by using the proposed method.

Radar HRR Profiles Recognition Based on SVM with Power-Transformed-Correlation Kernel

Radar automatic target recognition (RATR) based on high-range- resolution (HRR) profiles and support vector machine (SVM) classifier is concerned. The physical mechanism of a RATR performance improvement approach, namely, performing the power transformation to the original HRR signatures, is analyzed based on the properties of HRR profiles. And a novel kernel function, power transformed correlation (PTC) kernel, is designed subsequently for SVM classifiers. The classification performance of SVM and maximum correlation coefficient (MCC) classifier are evaluated based on the measured data.

PCA and kernel PCA for radar high range resolution profiles recognition

Radar high range resolution profile (HRRP) contains target structure information. It is shown to be a promising signature for radar automatic target recognition. As a method for data dimension reduction and feature extraction, principle component analysis (PCA) and kernel PCA have found wide applications in pattern recognition field. According to the characteristics of target pose sensitivity and shift sensitivity, a localized PCA and a modified KPCA are proposed for radar HRRP recognition. Also the methods for selecting the kernel basis vectors and handling the range-shift alignment are carefully addressed. Finally, support vector machine (SVM) classifier is used to evaluate the classification performance based on measured data. Experimental results show the proposed methods are effective and KPCA outperforms PCA.

Multipath signal resolving and time delay estimation for high range resolution radar

High range resolution radar target signature is a promising signature for radar automatic target recognition (ATR), and the multipath time-delay is used for height finding in high range resolution radar recently. We address multipath time-delay estimation under the scenario of closely spaced multipath for high range resolution radar, the target scattered signature can be resolved via the estimated channel parameters. By using the prior information of the typical multipath channel structure in radar application, a channel constrained MODE-RELAX algorithm is proposed to estimate the channel parameters. The resolved target signatures are used to evaluate the classification performance. Numerical results show that the proposed algorithm can estimate the time-delay very well for many cases, and the classification performance can be improved via the resolved signal.

Target recognition based on radial basis function network

A method of radar target recognition by range profiles is developed, based on the radial basis function network (RBFN). The problem of producing suitable patterns for recognition is discussed. Then a heuristic clustering algorithm for training RBFN is proposed. It is shown, from theoretical analysis and experimental results of rotating platform imaging based on experimental data acquired in a microwave anechoic chamber, that target recognition based on range profile is promising in application of ATR system.

An Efficient Kernel Optimization Method for Radar High-Resolution Range Profile Recognition

A kernel optimization method based on fusion kernel for high-resolution range profile (HRRP) is proposed in this paper. Based on the fusion of-norm and-norm Gaussian kernels, our method combines the different characteristics of them so that not only is the kernel function optimized but also the speckle fluctuations of HRRP are restrained. Then the proposed method is employed to optimize the kernel of kernel principle component analysis (KPCA) and the classification performance of extracted features is evaluated via support vector machines (SVMs) classifier. Finally, experimental results on the benchmark and radar-measured data sets are compared and analyzed to demonstrate the efficiency of our method.

An Efficient Kernel Optimization Method for High Range Resolution Profile Recognition

A kernel optimization based on fusion kernel for HRRP is proposed in this paper. Based on the fusion of the 1-norm and 2-norm Gaussian kernels, our method combines the different characteristics of them so that not only is the kernel function optimized but also the speckle fluctuations of HRRP are restrained. Then on the radar measured data the presented method is employed to the kernel optimization of KPCA and the classification performance of the extracted is evaluated via a SVM classifier. Finally, experiment results are compared and analyzed, which prove our method effective

Speeding up SVM in test phase: application to radar HRRP ATR

In this paper, a simple method is proposed to reduce the number of support vectors (SVs) in the decision function. Because in practice the embedded data just lie into a subspace of the kernel-induced space, F, we can search a set of basis vectors (BVs) to express all the SVs according to the geometrical structure, the number of which is less than that of SVs. The experimental results show that our method can reduce the run-time complexity in SVM with the preservation of machines generalization, especially for the data of large correlation coefficients among input samples.

A kernel optimization method based on the localized kernel fisher criterion

It is wildly recognized that whether the selected kernel matches the data controls the performance of kernel-based methods. Ideally it is expected that the data is linearly separable in the kernel induced feature space, therefore, Fisher linear discriminant criterion can be used as a kernel optimization rule. However, the data may not be linearly separable even after kernel transformation in many applications, a nonlinear classifier is preferred in this case, and obviously the Fisher criterion is not the best choice as a kernel optimization rule. Motivated by this issue, in this paper we present a novel kernel optimization method by maximizing the local class linear separability in kernel space to increase the local margins between embedded classes via localized kernel Fisher criterion, by which the classification performance of nonlinear classifier in the kernel induced feature space can be improved. Extensive experiments are carried out to evaluate the efficiency of the proposed method.