当前位置: X-MOL 学术Cartography and Geographic Information Science › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Applying forces to generate cartograms: a fast and flexible transformation framework
Cartography and Geographic Information Science ( IF 2.354 ) Pub Date : 2020-05-07 , DOI: 10.1080/15230406.2020.1745092
Shipeng Sun 1, 2
Affiliation  

ABSTRACT

Automatic production of contiguous area cartograms has become practical with computer algorithms, particularly applying forces to rubber-sheets and simulating diffusion processes. Reformulating the existing force-based and rubber-sheet methods, this article presents a fast and flexible force-based computational framework for general space transformation with solid physical and mathematical foundations. When being applied to cartogram production, this framework guarantees topological integrity, allows flexible force generation, achieves fast convergence, and avoids extreme shape deformation. Benchmarked against the recently published fast flow-based diffusion method using five datasets of various volume, compactness, and complexity, this force-based framework is faster, reduces more shape deformation, and can produce cartograms with distinctive slim or inflated styles. Additionally, the force-based framework is robust and can handle complicated datasets, whereas the flow-based method produces errors from them. Although the flow-based diffusion method often reduces cartogram size error more than the force-based (typical weighted mean error is below 0.1% versus 1%), both are outstanding with such low levels of error and their cartograms generally register little visual differences. Overall, the force-based transformation framework provides a fast, flexible, and robust alternative to the diffusion method for cartogram production.



中文翻译:

用力生成地图:快速灵活的转换框架

摘要

使用计算机算法自动生成连续区域的制图已变得实用,特别是将力施加到橡胶板上并模拟扩散过程。重新定义了现有的基于力和橡胶片的方法,本文提出了一种快速而灵活的基于力的计算框架,用于具有坚实的物理和数学基础的常规空间转换。当应用于制图时,该框架可确保拓扑完整性,允许灵活的力生成,实现快速收敛并避免极端的形状变形。相对于最近发布的基于快速流动的基于扩散的方法,该方法使用了五个具有不同体积,紧凑性和复杂性的数据集,该基于力的框架更快,减少了更多的形状变形,并可以生成具有显着苗条或夸张样式的制图。此外,基于力的框架很健壮,可以处理复杂的数据集,而基于流的方法会产生错误。尽管基于流的扩散方法通常会比基于力的方法减少制图大小的误差(典型的加权平均误差低于0.1%对1%),但两者都具有如此低的误差水平,并且它们的制图通常几乎没有视觉上的差异。总体而言,基于力的转换框架为制图的扩散方法提供了一种快速,灵活和强大的替代方法。尽管基于流的扩散方法通常会比基于力的方法减少制图大小的误差(典型的加权平均误差低于0.1%对1%),但两者都具有如此低的误差水平,并且它们的制图通常几乎没有视觉上的差异。总体而言,基于力的转换框架为制图的扩散方法提供了一种快速,灵活和强大的替代方法。尽管基于流的扩散方法通常会比基于力的方法减少制图大小的误差(典型的加权平均误差低于0.1%对1%),但两者都具有如此低的误差水平,并且它们的制图通常几乎没有视觉上的差异。总体而言,基于力的转换框架为制图的扩散方法提供了一种快速,灵活和强大的替代方法。

更新日期:2020-05-07
down
wechat
bug