基于街景数据的绿视率模型构建及其影响因子研究

以行人的视觉直观感受为出发点, 以匈牙利塞克什白堡为研究区, 基于街景数据, 深入研究植被信息提取方法, 针对传统像素级分类容易造成过度提取的现象, 提出一种面向对象的街景图像分类方法, 构建了基于街景数据的绿视率模型, 并分析总结了街景图像拍摄时的水平视角、垂直视角、水平方向范围等镜头参数对绿视率计算的影响。研究结果表明: 面向对象的分类方法提升了街景图像分类的精度和效率, 为绿视率计算模型提供了新的数据源和计算方法; 采集街景图像时, 增加水平视角和垂直视角、扩大水平方向范围能使绿视率计算结果更加真实地反映行人视觉感受。构建的绿视率计算模型能从行人角度为街道绿化的布局和空间结构优化提供依据, 可为城市绿地规划设计、居住区视觉生态设计等提供参考依据。

Construction of green view index model based on street view data and research on its influence factors

In order to solve the problem that pixel level classification is easy to cause over extraction,this paper proposes an object-oriented classification method of street view data and constructs a green view index model from the perspective of visual perception of pedestrians,by taking the city of Székesfehérvár in Hungary as study area.The influence of lens parameters such as horizontal angle,vertical angle and horizontal direction range on the calculation of green view index is analyzed and summarized.The results show that object-oriented classification method can improve classification accuracy and efficiency of street view data.It provides a new data source and calculation method for the calculation of green view index model.Increasing horizontal and vertical angle of view and expanding the range of horizontal direction can make the calculation results of green view index more truly reflect the visual feeling of pedestrians,when collecting street view images.The green view index model can provide basis for the layout and spatial structure optimization of street green space from the perspective of pedestrians.It can also provide reference for urban green space planning and design and visual ecological design of residential areas.