The third view of space states that space is neither lifeless nor neutral,but a living structure capable of being more living or less living.The living structure is defined as a physical and mathematical structure or ...The third view of space states that space is neither lifeless nor neutral,but a living structure capable of being more living or less living.The living structure is defined as a physical and mathematical structure or simply characterized by the recurring notion(or the inherent hierarchy) of far more small substructures than large ones.The more substructures the more living,and the higher hierarchy of the substructures the more living.This paper seeks to lay out a new kind of GeoInformatics on the notion of living structure and on the third view of space.The new GeoInformatics aims not only to better understand the forms and processes of everyday space but also-maybe more importantly-to make the space or the Earth’s surface living or more living.We introduce two fundamental laws of living structure:Tobler’s law on spatial dependence or homogeneity and scaling law on spatial interdependence or heterogeneity.We further argue that these two laws favor statistics over exactitude,because the statistics tend to make a structure more living than the exactitude.We present the concept of living structure through some working examples and make it clear how a living structure differs from a non-living or less-living structure.In order to make a structure or space living or more living,we introduce two design principles-differentiation and adaptation-using two paintings and two city plans as working examples.The new GeoInformatics is a science of living structure,dealing with a wide range of scales of the everyday space,from the smallest scale of ornaments on walls to the scale of the entire Earth’s surface.展开更多
This paper presents a free viewpoint video(FVV)system based on ray-space interpolation method.The new algorithm matches individual pixels in corresponding scanline pairs by using a dynamic program-ming technique.A spa...This paper presents a free viewpoint video(FVV)system based on ray-space interpolation method.The new algorithm matches individual pixels in corresponding scanline pairs by using a dynamic program-ming technique.A sparse intermediate view disparity map is projected from matched pixels firstly,andthe holes(occluded pixels)are filled in by propagating the disparity of neighboring background pixels.After interpolating dense view images,an arbitrary virtual view image can be easily rendered from thedense ray-space converted from these view images.The proposed method is evaluated on the Middleburydata set arid compared with other methods,experimental results show that the better quality of the inter-mediate view is obtained and the corresponding computational complexity is reduced significantly.展开更多
This paper explored relevant contents, methods and comprehensive efficiency of organic growth view in residential spaces based on personal experience, so as to fi gure out a new approach of residential space design wi...This paper explored relevant contents, methods and comprehensive efficiency of organic growth view in residential spaces based on personal experience, so as to fi gure out a new approach of residential space design with high economic efficiency, participation and sustainable connotations.展开更多
为了提高多视图深度估计结果精度,提出一种基于自适应空间特征增强的多视图深度估计算法。设计了由改进后的特征金字塔网络(feature pyramid network,FPN)和自适应空间特征增强(adaptive space feature enhancement,ASFE)组成的多尺度...为了提高多视图深度估计结果精度,提出一种基于自适应空间特征增强的多视图深度估计算法。设计了由改进后的特征金字塔网络(feature pyramid network,FPN)和自适应空间特征增强(adaptive space feature enhancement,ASFE)组成的多尺度特征提取模块,获取到具有全局上下文信息和位置信息的多尺度特征图像。通过残差学习网络对深度图进行优化,防止多次卷积操作出现重建边缘模糊的问题。通过分类的思想构建focal loss函数增强网络模型的判断能力。由实验结果可知,该算法在DTU(technical university of denmark)数据集上和CasMVSNet(Cascade MVSNet)算法相比,在整体精度误差、运行时间、显存资源占用上分别降低了14.08%、72.15%、4.62%。在Tanks and Temples数据集整体评价指标Mean上该模型优于其他算法,证明提出的基于自适应空间特征增强的多视图深度估计算法的有效性。展开更多
基金Swedish Research Council FORMAS through the ALEXANDER project (No.2017-00824)。
文摘The third view of space states that space is neither lifeless nor neutral,but a living structure capable of being more living or less living.The living structure is defined as a physical and mathematical structure or simply characterized by the recurring notion(or the inherent hierarchy) of far more small substructures than large ones.The more substructures the more living,and the higher hierarchy of the substructures the more living.This paper seeks to lay out a new kind of GeoInformatics on the notion of living structure and on the third view of space.The new GeoInformatics aims not only to better understand the forms and processes of everyday space but also-maybe more importantly-to make the space or the Earth’s surface living or more living.We introduce two fundamental laws of living structure:Tobler’s law on spatial dependence or homogeneity and scaling law on spatial interdependence or heterogeneity.We further argue that these two laws favor statistics over exactitude,because the statistics tend to make a structure more living than the exactitude.We present the concept of living structure through some working examples and make it clear how a living structure differs from a non-living or less-living structure.In order to make a structure or space living or more living,we introduce two design principles-differentiation and adaptation-using two paintings and two city plans as working examples.The new GeoInformatics is a science of living structure,dealing with a wide range of scales of the everyday space,from the smallest scale of ornaments on walls to the scale of the entire Earth’s surface.
基金the Natural Science Foundation of China(No.60472100,60672073)the Program for New Century Excellent Talents in University(No.NCET-06-0537)the Key Project of Chinese Ministry of Education(No.206059)
文摘This paper presents a free viewpoint video(FVV)system based on ray-space interpolation method.The new algorithm matches individual pixels in corresponding scanline pairs by using a dynamic program-ming technique.A sparse intermediate view disparity map is projected from matched pixels firstly,andthe holes(occluded pixels)are filled in by propagating the disparity of neighboring background pixels.After interpolating dense view images,an arbitrary virtual view image can be easily rendered from thedense ray-space converted from these view images.The proposed method is evaluated on the Middleburydata set arid compared with other methods,experimental results show that the better quality of the inter-mediate view is obtained and the corresponding computational complexity is reduced significantly.
文摘This paper explored relevant contents, methods and comprehensive efficiency of organic growth view in residential spaces based on personal experience, so as to fi gure out a new approach of residential space design with high economic efficiency, participation and sustainable connotations.
文摘为了提高多视图深度估计结果精度,提出一种基于自适应空间特征增强的多视图深度估计算法。设计了由改进后的特征金字塔网络(feature pyramid network,FPN)和自适应空间特征增强(adaptive space feature enhancement,ASFE)组成的多尺度特征提取模块,获取到具有全局上下文信息和位置信息的多尺度特征图像。通过残差学习网络对深度图进行优化,防止多次卷积操作出现重建边缘模糊的问题。通过分类的思想构建focal loss函数增强网络模型的判断能力。由实验结果可知,该算法在DTU(technical university of denmark)数据集上和CasMVSNet(Cascade MVSNet)算法相比,在整体精度误差、运行时间、显存资源占用上分别降低了14.08%、72.15%、4.62%。在Tanks and Temples数据集整体评价指标Mean上该模型优于其他算法,证明提出的基于自适应空间特征增强的多视图深度估计算法的有效性。
