基于长波的红外偏振成像技术研究

发布时间：2018-05-23 14:04

本文选题：红外成像 + 红外偏振成像　；参考：《天津大学》2016年硕士论文

【摘要】：目前红外探测被广泛应用,但传统红外探测设备在雾霾、烟尘、伪装等复杂环境下已经无法的探测需求。经研究和试验表明,红外偏振成像技术能够有效解决上述问题,因此红外的偏振成像技术研究具有重要意义和广阔的应用前景。红外偏振成像技术的研究主要是针对红外的偏振成像特性和成像应用的研究。本文以长波红外偏振成像为研究对象进行分析和研究红外的偏振成像特性和成像应用。本文基于红外成像应用系统,根据红外偏振成像生成原理,制定了红外偏振成像方案,搭建了长波红外偏振成像系统;利用长波红外偏振成像系统平台,采样景物的红外偏振信息,获得景物的红外信息和偏振信息;然后根据红外的偏振成像特点,选取图像融合算法,运用Matlab工具进行图像仿真处理,最后根据融合图像仿真结果研究长波红外的偏振特性。为了更好的应用红外的偏振信息,本文选取了相应的图像融合算法,具体步骤如下:首先采用双边滤波法对图像进行处理,去除噪声,增强物体边缘信息;然后基于多尺度的Retinex理论对融合算法进行了改进,增强了图像的同时,尽量排除外界干扰,还原物体本来面目;再通过加权平均融合规则对图像进行融合,将图像信息从对数域映射到显示域;最后通过非线性灰度变换,进一步凸显图像细节信息,加强了视觉舒适度。通过对仿真融合图像的评价,可以验证该融合算法在凸显图像细节和提升图像对比度方面效果突出,能够较好满足红外偏振成像系统的需要。红外偏振融合图像仿真结果表明:与红外探测本身相比,红外的偏振信息在金属探测和防伪探测上具有较强的探测能力,能够探测到金属目标的更多细节,能够透过防伪网探测到更多目标信息,但红外的偏振信息也易受水雾的干扰。本文基于长波的红外偏振成像技术研究,验证了红外偏振成像系统处理流程的可执行性,探测和发现红外偏振成像特点,为后续红外偏振成像系统的实时应用和工程化应用等研究打好基础。
[Abstract]:At present, infrared detection is widely used, but the traditional infrared detection equipment in haze, smoke, camouflage and other complex environments has been unable to detect the requirements. The research and experiments show that the infrared polarization imaging technology can effectively solve the above problems, so the study of infrared polarization imaging technology has important significance and broad application prospects. The research of infrared polarization imaging is mainly focused on the characteristics and applications of infrared polarization imaging. In this paper, the polarization imaging characteristics and imaging applications of long wave infrared polarization imaging are analyzed and studied. Based on the infrared imaging application system, according to the principle of infrared polarization imaging generation, this paper formulates the infrared polarization imaging scheme, builds the long-wave infrared polarization imaging system, and makes use of the long-wave infrared polarization imaging system platform. The infrared polarization information of the scene is sampled, and the infrared information and polarization information of the scene are obtained. Then, according to the characteristics of the infrared polarization imaging, the image fusion algorithm is selected, and the image simulation processing is carried out by using the Matlab tool. Finally, the polarization characteristics of long wave infrared are studied according to the fusion image simulation results. In order to better use the infrared polarization information, this paper selects the corresponding image fusion algorithm, the specific steps are as follows: firstly, the bilateral filtering method is used to process the image to remove noise and enhance the edge information; Then, the fusion algorithm is improved based on the multi-scale Retinex theory to enhance the image, eliminate the external interference, restore the original appearance of the object, and then fuse the image by the weighted average fusion rule. The image information is mapped from the logarithmic domain to the display field. Finally, the image details are further highlighted by nonlinear gray level transformation, and the visual comfort is enhanced. Through the evaluation of the simulated fusion image, it can be verified that the fusion algorithm can highlight the image details and enhance the image contrast, and can better meet the needs of infrared polarization imaging system. The simulation results of infrared polarization fusion image show that compared with infrared detection itself, infrared polarization information has stronger detection ability in metal detection and anti-counterfeiting detection, and can detect more details of metal target. More target information can be detected through the anti-counterfeiting network, but the infrared polarization information is also vulnerable to water mist interference. Based on the research of long wave infrared polarization imaging technology, this paper verifies the executable of infrared polarization imaging system, detects and discovers the characteristics of infrared polarization imaging. It lays a good foundation for the research of infrared polarization imaging system in real time and engineering.
【学位授予单位】：天津大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP391.41

【参考文献】