被动毫米波成像技术研究

【作者】 纪如霆；

【导师】 许建中；

【作者基本信息】 南京理工大学 ， 电路与系统， 2005， 硕士

【摘要】 毫米波是一种波长位于微波和红外光谱之间的电磁辐射,兼有红外和微波的共同特征,从而形成其特有的性能。被动毫米波成像是毫米波技术领域90年代以来的一个研究热点,目前被动毫米波成像普遍采用单天线波束机械扫描工作方式,该法存在着分辨率不高以及灵敏度较低等不足。近年来,随着毫米波固体器件和信号处理技术的快速发展,被动毫米波高分辨率成像技术发展十分迅速,八十年代末出现的毫米波焦平面多波束阵列成像技术使成像精度、实时性大大提高,并逐渐开始应用于军事和民用领域。 本文首先总结了被动毫米波辐射成像的一般理论和原理,并对部分技术指标进行了分析;结合教研室的毫米波单波束扫描成像实验,对实验数据进行提取处理成像,验证扫描成像的可行性;其后由此引入毫米波焦平面阵列成像技术的分析,从数据率、多波束、空间分辨率和空间采样特性入手,并结合两种成像接收机和三种扫描方式的特性等多个方面综合分析;重点研究了被动毫米波图像处理的问题,根据毫米波图像空间分辨率低和图像模糊的特点,探讨了非线性外推的高分辨率的图像复原增强算法及其改进算法,和对基于多幅的超分辨率图像处理凸集投影算法(POCS)进行了重点的研究;最后研究和设计了线阵并扫的DSP实时图像处理平台。更多还原

【Abstract】 Millimeter-wave (MMW) wavelengths is at the field between infrared and microwave, so it has the same characteristics of two ranges. Passive millimeter-wave imaging has been a hot research from 90s’ . Problems of current imaging systems, which relied on mechanical scanning of a single beam across the scene, are low spatial resolution and poor sensitivity. In recent years, the technology of passive millimeter-wave imaging of high spatial resolution develops very fast depending on the rapidly development of solid state components and signal processing technology. Focal plane array, multi-beam array imaging systems, which promise an increase in resolution and real-time processing, is gradually used in both the military and civilian applications.Firstly, we explain the common theory and principles of passive millimeter-wave imaging and analyze several technical indexes of the imaging system. Secondly, we got the passive MMW image by processing the experiment data and verify the possibility of the single channel scanning passive MMW imaging. The technology of focal plane arrays is introduced, and the features of data rate, multi-beam, spatial resolution and spatial sampling are analyzed in detail, and the features of two kinds receiver and three scanning methods are presented. Thirdly, this thesis focuses upon passive millimeter-wave image processing. Depending on millimeter-wave characteristics of low spatial resolution and image blurring, we discuss the nonlinear extrapolation algorithm, reformative algorithm and a super-resolution image reconstruction algorithm based on projection onto convex sets (POCS). At last we design multi-channel parallel scanning video image real-time processor based on DSP in this paper.更多还原