基于毫米波三维成像的机场人员安检关键技术研究

In recent years, affected by the turmoil of the world, security of airports and other public places increasingly becomes the focus of governments. But as weapons equipped by terrorists becomes more and more advanced in recent years, the traditional testing methods showed their shortage, which are difficult to detect illegal items and weapons made of new materials in complex environment. As a new means of high resolution imaging, millimeter-wave imaging technology has very strong transmission ability, which is able to penetrate clothes and packaging materials to detect ceramic knives and plastic pistol. Compared with X ray, millimeter-wave radiation energy is very small, which will not produce ionization damage to objects illuminated, thus can be used to realize non-contact, non-destructive detection, which is very suitable for airport security, thus has important research significance. This study started from the millimeter wave frequency target echo characteristics, combined with target the conditions needed for 3-D imaging and resolution constraint, millimeter-wave three-dimensional imaging model is set up. Combining with the Circular-SAR imaging algorithm, interferometric imaging methods and the real-time demand required by security inspection, we put forward millimeter wave three-dimensional imaging algorithm suitable for of airport security. On this basis, we propose the automatic detection and recognition algorithm based on 3-D imaging data to implement rapidly and effectively concealed detection.

近年来，受动荡的世界格局影响，机场等公共场所的安全越来越成为各国政府关注的焦点。然而由于近年来恐怖分子配备的武器越来越先进，传统的检测手段显露出了不足，难以快速检测识别复杂环境中的非法物品以及由新兴材料制作的非法武器。毫米波成像技术作为一种新兴的高分辨率成像手段，具有很强的透射能力，能够穿透衣服和包装材料，可以探测陶瓷匕首、塑料手枪等武器。相比于X射线，毫米波辐射能量很小，不会对所照射物体产生电离破坏作用，可用来实现非接触、非破坏性的探测，非常适用于机场安检，因而具有重大研究意义。本课题从毫米波频段下目标回波特性出发，研究目标三维成像所需条件及目标分辨率约束原因，建立毫米波三维成像模型。结合圆周SAR成像算法、干涉成像方法以及安检所要求的实时性要求，提出适用于机场安检的毫米波三维快速成像算法。在此基础上，提出基于三维成像数据的自动检测识别算法，实现对人体隐匿物品的快速有效检测。

近年来，受动荡的世界格局影响，机场等公共场所的安全越来越成为各国政府关注的焦点。然而由于近年来恐怖分子配备的武器越来越先进，传统的检测手段显露出了不足，难以快速检测识别复杂环境中的非法物品以及由新兴材料制作的非法武器。毫米波成像技术作为一种新兴的高分辨率成像手段，具有很强的透射能力，能够穿透衣服和包装材料，可以探测陶瓷匕首、塑料手枪等武器。相比于X射线，毫米波辐射能量很小，不会对所照射物体产生电离破坏作用，可用来实现非接触、非破坏性的探测，非常适用于机场安检，因而具有重大研究意义。本课题基于圆周SAR成像机理与数学模型，开展了有关毫米波近场三维SAR快速成像方法研究，并结合目标在三维雷达图像中的散射特征分析了相关积累角变化与成像分辨率及聚焦深度的关系，提出了基于子孔径的毫米波三维快速成像算法、子孔径分割与拼接方法以及基于三维数据的异质目标自动检测识别算法，通过仿真实验结果讨论了算法性能，并搭建了毫米波近场三维成像平台进行了验证实验，实现了毫米波三维成像及隐匿物品的自动检测识别。