基于ISAR图像的舰船目标特征提取方法研究

发布时间：2018-07-21 11:31

【摘要】：逆合成孔径雷达(ISAR)作为一种高分辨的成像雷达,能够全天候和全天时的进行远距离的目标成像,在军用和民用领域具有非常重要的价值。随着成像技术的不断提高,ISAR图像的清晰度越来越高,ISAR已广泛应用于空天目标和海面目标成像,同时ISAR图像包含目标精细的几何结构、散射点分布等多种信息,对目标的判别有重要的价值,因此如何从ISAR图像中获取目标的信息进行目标的识别分类已经成为研究的热门领域。本文以此为出发点,研究了基于ISAR图像的舰船目标的特征提取方法,主要工作概括如下:1.介绍ISAR的成像原理以及ISAR的信号处理方法,包括平动补偿技术和成像算法,分析了图像中存在的质量问题,并介绍了如何利用电磁仿真软件FEKO进行数据仿真,最后进行了实验验证。2.对ISAR图像的质量评估方法进行了深入的研究,通过分析图像质量存在的问题,例如存在斑点噪声、条纹干扰、重影、运动模糊等问题,提出了一种图像质量分类的方法,通过提取能够反映图像质量信息的特征对图像进行分类,并对该方法进行了验证。3.研究了基于ISAR图像舰船目标的长度特征和属性散射中心特征的提取方法。(1)针对大小船的分类应用,通过中心线的提取能够对目标进行有效的分类,由于常用的Hough变换等方法在提取目标的长度时存在较大的误差,本文利用PCA变换进行目标长度的提取,该方法利用降维的方法找出目标在主轴线上的投影。(2)为了对舰船目标进行更细致的分类,研究了属性散射中心特征,本文利用正交匹配追踪算法提取目标的属性散射中心,并进行了实验验证相关算法的有效性。本文最后对全文的工作进行了总结,并指出有待进一步研究的问题。
[Abstract]:Inverse synthetic Aperture Radar (ISAR), as a high-resolution imaging radar, can be used for long-range target imaging all weather and all day. It is of great value in military and civil fields. With the improvement of imaging technology, the sharpness of ISAR image has been widely used in space-sky target and sea surface target imaging, and ISAR image contains a variety of information, such as the precise geometric structure of the target, the distribution of scattering points, and so on. So how to obtain the target information from ISAR image for target recognition and classification has become a hot research field. In this paper, the feature extraction method of ship target based on ISAR image is studied. The main work is summarized as follows: 1. This paper introduces the imaging principle of ISAR and the signal processing method of ISAR, including the translation compensation technology and imaging algorithm, analyzes the quality problems existing in the image, and introduces how to use the electromagnetic simulation software FEKO to simulate the data. Finally, the experiment was carried out to verify. 2. The quality evaluation method of ISAR image is studied deeply. A method of image quality classification is proposed by analyzing the problems of image quality, such as speckle noise, fringe interference, double shadow, motion blur and so on. The image is classified by extracting the features which can reflect the image quality information, and the method is verified. 3. The extraction method of ship target length feature and attribute scattering center feature based on ISAR image is studied. (1) according to the classification application of large and small ships, the target can be effectively classified by centerline extraction. Due to the error of Hough transform in extracting the length of the target, the PCA transform is used to extract the length of the target. This method uses dimensionality reduction method to find out the projection of the target on the main axis. (2) in order to classify the ship targets more carefully and study the feature of the attribute scattering center, the orthogonal matching tracking algorithm is used to extract the attribute scattering center of the target. Experiments are carried out to verify the effectiveness of the algorithm. Finally, this paper summarizes the work of the paper and points out the problems to be further studied.
【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】：E925;TN957.52

【相似文献】