光学等高压缩三维成像方法研究

项目名称： 光学等高压缩三维成像方法研究

项目编号： No.61271332

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： 顾国华

作者单位： 南京理工大学

项目金额： 85万元

中文摘要： 本项目针对光学三维成像的应用和技术需求，提出等高压缩的概念和原理，围绕涉及三维信息稀疏测量的科学问题：等高信息矩阵设计、间接观测量构造与等高信息矩阵优化估计，开展光学等高压缩三维成像方法研究。分析待测目标三维形状特征对发射光学信号的影响，建立微小平面的信号传递模型，研究正交距离基的设计方法并探索其参数估计方法，构造与待测目标三维形状特征等价的可稀疏等高信息矩阵；研究等高信息矩阵的稀疏采样方式以及测量矩阵的物理实现途径，抽取直接观测量的时频域特征，探索其与等高信息矩阵之间的关联模式及量化描述手段，构造等高信息矩阵的间接观测量；研究等高信息矩阵约束条件变换方法，设计稀疏度非相干的快速贪婪算法，实现等高信息的最优化估计，完成待测目标三维信息精确高效重构。本项目是压缩感知思想的延伸和发展，研究成果对于丰富压缩感知理论体系的内涵，理论研究和应用、推动光学三维成像技术发展等方面都具有重要的作用。

中文关键词： 光学三维成像；压缩感知；计算成像；小波；

英文摘要： We aim to develop new method of optical imaging based on the new concept of contour compressed. We start by beginning with constructing contour information matrix, acquiring the indirect measurement values, reconstructing the contour information matrix. We analyze the transmitted optical signal influenced by 3D shape characteristics of the object, build the underwater optical signal transmission model based on small facet model, research on the methods of defining the orthogonal basis of depth and estimating parameters, and construct sparse contour information matrix of 3D shape characteristics of the object. We research on sparsely sampling of contour information matrix and physical realization of the measurement matrix, acquire the characteristics of received signal in time and frequency domain, search for the connection pattern and quantitative description between characteristics and contour information matrix, and construct the indirect measurement variable representing contour information matrix. We research on the method of changing constraint condition about contour information matrix, design rapid greedy algorithm which is incoherent with sparsity, realize the optimization of contour information and reconstruct 3D information of the object accurately and efficiently. Our project is an extension and devel

英文关键词： optical three-dimensional imaging；compressed sensing；computational imaging；wavelets；