基于降维技术的大比例尺城市居民地匹配方法研究
发布时间：2019-04-18 07:59
【摘要】：随着社会的不断发展，，人们赖以生存的地理空间持续发生着变化，城市居民地作为空间目标变化中最为活跃的要素之一，是空间数据更新的主要内容。面状居民地因其分布密集、形态复杂、轮廓不规则等特点，已成为空间数据匹配中最为复杂最具挑战性的部分。 综合分析了现有面要素常见匹配方法，本文提出了采用降维技术对大比例尺城市面状居民地进行匹配的新方法。主要工作及创新点体现在以下几个方面： （1）提出了采用降维技术进行面状居民地匹配的新方法。对面状居民地进行降维处理，把面状居民地转化为能够反映其轮廓主要形态特征的骨架线，一方面有效降低了匹配的复杂性及计算量；另一方面线要素的许多匹配算法可引入进来，有效拓展了面要素匹配的方法与技术途径。 （2）研究了面状居民地骨架线提取与化简方法。分析了常见面要素骨架线提取算法，根据面状居民地自身特点，对居民地骨架线提取算法进行了统一；在分析常见线要素化简方法的基础上，提出了基于自身几何特征的骨架线化简算法。 （3）构建了基于骨架线的居民地匹配方法。在综合分析已有线要素匹配算法的基础上，梯次提出了基于骨架线缓冲区法、骨架线积分面积差值率法、骨架线傅立叶变换法等三种居民地匹配方法，并对三种方法的可行性和科学性进行了对比验证。 （4）提出基于约束城市骨架线网的居民地匹配方法。面状居民地因其在空间分布上的离散性，导致对其整体结构与关系描述较为困难。而作为面状居民地补集的城市空白区域，其分布是连续的，空白区域能够较好地反映居民地及其相互之间的几何与方位等关系。据此，本文提出了基于约束城市骨架线网的居民地匹配方法，在城市道路网的约束下，提取城市空白区域骨架线，通过对约束城市骨架线网进行匹配，达到对居民地匹配的目的。 （5）实现了基于降维技术的面状居民地匹配实验系统。以同一区域内的多源大比例尺城市面状居民地数据集为实验对象，对本文理论与方法的科学性进行了验证。
[Abstract]:With the continuous development of society, the geographical space on which people depend for survival is constantly changing. As one of the most active elements in the change of spatial objectives, urban residential land is the main content of spatial data updating. Because of its dense distribution, complex shape and irregular contour, the surface residential area has become the most complex and challenging part of spatial data matching. Based on a comprehensive analysis of the common matching methods of surface elements, this paper presents a new method of matching large-scale urban surface residents with dimensionality reduction technique. The main work and innovation points are as follows: (1) A new method of surface resident land matching based on dimension reduction technology is proposed. On the one hand, the complexity and calculation of matching can be effectively reduced by reducing the dimension of the opposite residents and transforming them into skeleton lines which can reflect the main morphological features of the contours. On the other hand, many matching algorithms of line features can be introduced, which effectively extends the methods and technical approaches of surface feature matching. (2) the method of extracting and simplifying the skeleton line of the area resident is studied. This paper analyzes the algorithm of extracting skeleton line of common meet feature, and unifies the algorithm of extracting skeleton line of resident ground according to the characteristics of area resident land. Based on the analysis of common line feature reduction methods, a skeleton line reduction algorithm based on its own geometric features is proposed. (3) the method of resident-land matching based on skeleton line is constructed. On the basis of comprehensive analysis of the matching algorithm of wired elements, three resident matching methods based on skeleton line buffer method, skeleton line integral area difference ratio method and skeleton line Fourier transform method are proposed, which are based on skeleton line buffer method, skeleton line integral area difference ratio method and skeleton line Fourier transform method. The feasibility and scientificalness of the three methods are compared and verified. (4) A method of resident-land matching based on constrained urban skeleton network is proposed. Because of its discreteness in spatial distribution, it is difficult to describe its whole structure and relationship. The distribution of urban blank area is continuous, and the blank area can better reflect the relationship between geometry and orientation of residential areas. According to this, this paper puts forward a method of resident-land matching based on constrained urban skeleton line network. Under the constraint of urban road network, the urban blank area skeleton line is extracted, and the constrained urban skeleton line network is matched. To achieve the purpose of matching the residents. (5) the surface resident matching experiment system based on dimension reduction technology is realized. Taking the multi-source and large-scale urban residential land data set in the same area as the experimental object, the scientific nature of the theory and method in this paper is verified.
【学位授予单位】：解放军信息工程大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：P208
本文编号：2459883
[Abstract]:With the continuous development of society, the geographical space on which people depend for survival is constantly changing. As one of the most active elements in the change of spatial objectives, urban residential land is the main content of spatial data updating. Because of its dense distribution, complex shape and irregular contour, the surface residential area has become the most complex and challenging part of spatial data matching. Based on a comprehensive analysis of the common matching methods of surface elements, this paper presents a new method of matching large-scale urban surface residents with dimensionality reduction technique. The main work and innovation points are as follows: (1) A new method of surface resident land matching based on dimension reduction technology is proposed. On the one hand, the complexity and calculation of matching can be effectively reduced by reducing the dimension of the opposite residents and transforming them into skeleton lines which can reflect the main morphological features of the contours. On the other hand, many matching algorithms of line features can be introduced, which effectively extends the methods and technical approaches of surface feature matching. (2) the method of extracting and simplifying the skeleton line of the area resident is studied. This paper analyzes the algorithm of extracting skeleton line of common meet feature, and unifies the algorithm of extracting skeleton line of resident ground according to the characteristics of area resident land. Based on the analysis of common line feature reduction methods, a skeleton line reduction algorithm based on its own geometric features is proposed. (3) the method of resident-land matching based on skeleton line is constructed. On the basis of comprehensive analysis of the matching algorithm of wired elements, three resident matching methods based on skeleton line buffer method, skeleton line integral area difference ratio method and skeleton line Fourier transform method are proposed, which are based on skeleton line buffer method, skeleton line integral area difference ratio method and skeleton line Fourier transform method. The feasibility and scientificalness of the three methods are compared and verified. (4) A method of resident-land matching based on constrained urban skeleton network is proposed. Because of its discreteness in spatial distribution, it is difficult to describe its whole structure and relationship. The distribution of urban blank area is continuous, and the blank area can better reflect the relationship between geometry and orientation of residential areas. According to this, this paper puts forward a method of resident-land matching based on constrained urban skeleton line network. Under the constraint of urban road network, the urban blank area skeleton line is extracted, and the constrained urban skeleton line network is matched. To achieve the purpose of matching the residents. (5) the surface resident matching experiment system based on dimension reduction technology is realized. Taking the multi-source and large-scale urban residential land data set in the same area as the experimental object, the scientific nature of the theory and method in this paper is verified.
【学位授予单位】：解放军信息工程大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：P208
【参考文献】
相关期刊论文 前1条
1 郭邦梅;王涛;赵荣;梁勇;;面状要素骨架线提取算法的研究[J];测绘通报;2010年12期
相关博士学位论文 前1条
1 郭黎;多源地理空间矢量数据融合理论与方法研究[D];解放军信息工程大学;2008年
本文编号：2459883
本文链接：https://www.wllwen.com/kejilunwen/dizhicehuilunwen/2459883.html
