多物理场图像处理与应用

项目名称： 多物理场图像处理与应用

项目编号： No.61271015

项目类型： 面上项目

立项/批准年度： 2013

项目学科： 无线电电子学、电信技术

项目作者： 岳军

作者单位： 青岛理工大学

项目金额： 82万元

中文摘要： 干涉数据（干涉图）处理是各种干涉成像系统中的关键技术之一。针对偏微分方程（PDE）方法在干涉成像中存在的参数估计和实时处理等问题，本课题在PDE图像处理方法的基础上，以干涉合成孔径声呐 (InSAS）成像系统为测试平台，提出一种多物理场图像处理的新方法。该方法建立了新的图像处理模型和数学处理框架，以及快速有效的数值计算方法，将干涉图处理视为扩散场和渗流场等多个物理场的演化结果。在数学模型上表现为满足初边值条件下多个偏微分方程的耦合，图像处理结果是演化方程组在图像空间上的整体弱解；理论上解决解的适定性和多场融合的快速算法；技术上依据InSAS系统中相位展开等要素确定演化时间。该方法具有直观的物理意义和严格的数学机制，避免了PDE方法中的参数估计，提高了算法的实效性，拓展了PDE方法的理论研究和应用实践。成果有望应用于我国自主研制的InSAS系统，提高成像中的高程精度，对相关研究提供借鉴。

中文关键词： 多物理场；图像处理；非线性偏微分方程；耦合算法；InSAS干涉图

英文摘要： The problem of interferogram processing is one of the key technology in interference imaging systems. The method of PDE in image processing has the problem of parameter estimation and real-time processing in interference imaging. For the problem above, a new method of Multiphysics image processing is proposed, which will be researched on the foundation of PDE image processing and tested on the platform of the InSAS(Interferometric Synthetic Aperture Sonar) imaging system. A new image processing model is established as well as the mathematical theory frame and fast efficient numerical computation method. The program processing is viewed as the multi-physical evolution result such as the field of the diffusion and the seepage. The performance in mathematics is the coupled PDEs satisfying the initial and boundary value conditions. The result of image processing is the global weak solution of the evolution equations in the image space. In theory, it should be solved for the well-posed solution and the fast algorithm of the multi-physical evolution. In technology, the evolution time should be set based on the require of phase unwrapping in the InSAS system. This method has the visualized physical significance and serious mathematical theory, which can avoid the parameter estimation of the PDE method. The efficiency o

英文关键词： Multiphysics；Image processing；Nonlinear PDE；Coupled algorithm；InSAS interferogram