海洋环境数据场建模与可视化技术研究

发布时间：2018-10-24 18:46

【摘要】：随着海洋在政治、经济、军事领域的战略意义日渐凸显,人们对海洋的探索和研究呈现蓬勃发展之势。谁能够掌握海洋的客观规律,充分利用海洋资源,谁就占据了发展的先机。海洋环境数据分析是海洋探索的重要手段,为海洋环境数据建立符合海洋特点的高效数据模型,是充分了解、利用海洋信息的前提;同时,对海洋环境数据的可视化,使抽象数据直观具体,将有利于人们掌握海洋的本质和内涵。论文围绕海洋环境数据场的建模、矢量场数据插值、矢量场可视化三方面内容展开研究,针对存在的问题,提出了相应的解决方法和思路,论文的主要工作如下:(1)在数据场建模方面,提出了基于时空格网二维基态修正模型的海洋环境数据场建模方法。首先分析了海洋环境数据的特点,针对现有模型的不足,提出了更符合海洋特点规律的海洋环境时空格网二维基态修正模型。该模型以同一年初始月数据为基态数据,以差文件的形式存储其他月份数据,形成“基态年不同月份”数据维度;然后以该维度中各月份数据为基础,对不同年的相应月份再次记录差文件,构建“不同年相同月份”数据维度,从而形成二维基态修正模型。最后基于此模型,实现了海洋环境数据建模。提出的模型较大限度地减少了数据冗余,更有利于海洋环境数据的分析。(2)在矢量场插值方面,基于流体的运动规律和特点,提出了基于拉格朗日流体质点运动模型的矢量场数据插值算法。算法首先对所研究区域内的流体进行拉格朗日质点建模;然后选取距离待插值点最近的六个采样点,基于拉格朗日模型估算其下一时刻的位置;利用采样点前后两个时刻的位置坐标,解算出区域流体运动方程的待定系数;最后根据运动方程确定待插值点运动后的位置坐标,进而反推出任一点矢量数据。算法采用二次多项式对区域流体运动进行拟合,更符合区域流体的运动规律,所得插值结果更为精确。(3)在矢量场可视化方面,分析了经典线积分卷积及其改进算法在表达矢量场强度方面存在的不足,提出了一种基于场强驱动的纹理可视化算法。算法首先将矢量场强度与输入噪声纹理结合,生成场强驱动的输入噪声纹理;然后利用线积分卷积得到灰度输出纹理;最后基于伪彩色增强映射出既能展现矢量场方向,又能反映矢量大小的彩色输出纹理图像。该算法所得到的彩色输出纹理能够清晰、明确地展现矢量场的全方位信息。
[Abstract]:With the increasing strategic significance of the ocean in the political, economic and military fields, people's exploration and research on the ocean are showing a trend of vigorous development. Who can grasp the objective law of the ocean and make full use of the marine resources, who will occupy the first opportunity of development. The analysis of marine environmental data is an important means of ocean exploration. It is a prerequisite to fully understand and utilize marine information to establish an efficient data model for marine environmental data, and at the same time, to visualize marine environmental data. Making abstract data intuitionistic and concrete will help people grasp the essence and connotation of ocean. This paper focuses on the modeling of marine environment data field, vector field data interpolation, vector field visualization, and puts forward the corresponding solutions and ideas for the existing problems. The main work of this paper is as follows: (1) in the aspect of data field modeling, a method of ocean environment data field modeling based on spatio-temporal grid two-dimensional ground state correction model is proposed. Firstly, the characteristics of marine environmental data are analyzed, and a two-dimensional ground state correction model based on space net is proposed, which is more consistent with the characteristics of marine environment, aiming at the shortcomings of the existing models. The model takes the monthly data of the beginning of the same year as the ground state data, stores the other month data in the form of difference files, and forms the data dimension of "different months of the year of the ground state", and then based on the data of each month in this dimension. The difference file is recorded again for the corresponding month of different years, and the data dimension of "different month of different years" is constructed, and a two-dimensional ground state correction model is formed. Finally, based on this model, the data modeling of marine environment is realized. The proposed model greatly reduces the data redundancy, and is more conducive to the analysis of marine environmental data. (2) in terms of vector field interpolation, based on the laws and characteristics of fluid motion, A vector field data interpolation algorithm based on Lagrange fluid particle motion model is proposed. Firstly, the Lagrange particle model is used to model the fluid in the studied region, and then the six sampling points closest to the interpolated point are selected to estimate the position of the next moment based on the Lagrange model. The undetermined coefficients of the regional fluid motion equations are solved by using the position coordinates of the two times before and after the sampling points, and the position coordinates after the motion of the points to be interpolated are determined according to the motion equations, and then the point vector data are deduced back. The quadratic polynomial is used to fit the regional fluid motion, and the interpolation results are more accurate. (3) in the aspect of vector field visualization, In this paper, the shortcomings of classical line integral convolution and its improved algorithm in expressing vector field intensity are analyzed, and a texture visualization algorithm based on field intensity is proposed. First, the vector field intensity is combined with the input noise texture to generate the input noise texture driven by the field intensity; then the gray output texture is obtained by linear integral convolution. Finally, the vector field direction can be displayed based on pseudo-color enhancement mapping. Also can reflect the vector size of the color output texture image. The color output texture obtained by this algorithm can clearly and clearly display the omnidirectional information of the vector field.
【学位授予单位】：国防科学技术大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：P717

【相似文献】