雷达高分辨距离像的散射点匹配识别和系统设计
发布时间：2018-08-24 11:15
【摘要】：基于高分辨距离像(HRRP)的目标识别一直是雷达自动目标识别领域的研究热点,此前已有一些经典的识别算法提出。基于目标散射点模型,HRRP是目标散射点子回波在径向距离维的投影向量和,对于飞机类目标,由于强散射点具有稀疏分布的特点,其散射特性也主要由强散射点子回波决定,如果能将强散射点提取出来,将有利于目标识别。然而,当前利用HRRP的散射点进行匹配识别的算法较少。本文的主要研究内容是在目标散射点模型和稀疏表示假设条件下,提取HRRP样本中散射点的强度和位置信息,然后采用点匹配算法进行分类识别,并在此基础上实现了HRRP实时识别系统。本文的研究内容概括如下:1.从雷达目标散射点模型出发推导得出高分辨距离像回波的数学表达式,在稀疏表示的假设条件下利用正交匹配追踪算法实现HRRP样本的散射点提取和噪声抑制。2.介绍了Hausdorff距离的基本定义,对提取的目标HRRP样本散射点集,采用基于Hausdorff距离的点匹配算法进行分类识别,并进一步提出一种噪声稳健的散射点匹配算法。基于实测数据的识别实验证明了该算法的有效性。3.实现雷达目标识别系统的设计开发,该系统由主机和信号处理板组成,本文主要研究两部分内容:1)主机端界面软件开发,包括软件开发环境的介绍,流程和框架设计,并在VC6.0编译器中采用MFC和OpenGL实现;2)信号处理板上的DSP识别程序,包括DSP程序开发的硬件资源介绍,算法实现以及程序优化,最后在Visual DSP++编译器中采用C语言实现。
[Abstract]:Target recognition based on high resolution range profile (HRRP) has been a hot topic in the field of radar automatic target recognition. Based on the scattering point model, HRRP is the projection vector sum of the radial distance dimension of the target scattering point echo. For aircraft targets, the scattering characteristics are mainly determined by the strong scattering point echo because of the sparse distribution of the strong scattering points. If strong scattering points can be extracted, it will be advantageous to target recognition. However, there are few algorithms to match and identify the scattering points of HRRP. The main research content of this paper is to extract the intensity and position information of scattering points in HRRP samples under the condition of target scattering point model and sparse representation hypothesis, and then use point matching algorithm to classify and identify. On this basis, the HRRP real-time recognition system is realized. The research contents of this paper are summarized as follows: 1. The mathematical expression of high resolution range echo is derived from the scattering point model of radar target. Under the assumption of sparse representation, orthogonal matching tracking algorithm is used to extract the scattering points of HRRP samples and to suppress the noise. In this paper, the basic definition of Hausdorff distance is introduced. The point matching algorithm based on Hausdorff distance is used to classify the scattered point set of target HRRP samples, and a noise robust scatter point matching algorithm is proposed. The recognition experiment based on the measured data proves the effectiveness of the algorithm. 3. 3. The design and development of radar target recognition system is realized. The system is composed of host computer and signal processing board. This paper mainly studies the software development of host interface in two parts: 1), including the introduction of software development environment, flow chart and frame design. The DSP recognition program on the signal processing board is implemented by MFC and OpenGL in the VC6.0 compiler, including the introduction of hardware resources, algorithm implementation and program optimization of DSP program development. Finally, C language is used in the Visual DSP compiler.
【学位授予单位】：西安电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN957.52
本文编号：2200650
[Abstract]:Target recognition based on high resolution range profile (HRRP) has been a hot topic in the field of radar automatic target recognition. Based on the scattering point model, HRRP is the projection vector sum of the radial distance dimension of the target scattering point echo. For aircraft targets, the scattering characteristics are mainly determined by the strong scattering point echo because of the sparse distribution of the strong scattering points. If strong scattering points can be extracted, it will be advantageous to target recognition. However, there are few algorithms to match and identify the scattering points of HRRP. The main research content of this paper is to extract the intensity and position information of scattering points in HRRP samples under the condition of target scattering point model and sparse representation hypothesis, and then use point matching algorithm to classify and identify. On this basis, the HRRP real-time recognition system is realized. The research contents of this paper are summarized as follows: 1. The mathematical expression of high resolution range echo is derived from the scattering point model of radar target. Under the assumption of sparse representation, orthogonal matching tracking algorithm is used to extract the scattering points of HRRP samples and to suppress the noise. In this paper, the basic definition of Hausdorff distance is introduced. The point matching algorithm based on Hausdorff distance is used to classify the scattered point set of target HRRP samples, and a noise robust scatter point matching algorithm is proposed. The recognition experiment based on the measured data proves the effectiveness of the algorithm. 3. 3. The design and development of radar target recognition system is realized. The system is composed of host computer and signal processing board. This paper mainly studies the software development of host interface in two parts: 1), including the introduction of software development environment, flow chart and frame design. The DSP recognition program on the signal processing board is implemented by MFC and OpenGL in the VC6.0 compiler, including the introduction of hardware resources, algorithm implementation and program optimization of DSP program development. Finally, C language is used in the Visual DSP compiler.
【学位授予单位】：西安电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN957.52
【参考文献】
相关硕士学位论文 前1条
1 张奔;基于高分辨距离像的目标识别算法和演示系统设计研究[D];西安电子科技大学;2011年
，本文编号：2200650
本文链接：https://www.wllwen.com/kejilunwen/wltx/2200650.html
