Vegetation Modeling And 3D Green Quantity Index Evaluation Based On L-system

Cities are important gathering places for human beings.The natural vegetation in the urban space not only affects the living environment of people and determines the quality of people’s life,but also plays a key role in regulating the whole regional ecosystem;and due to the complex 3D structure of plants and the numerous amounts of plants present in cities,it has been a worthwhile research topic for quantifying the amount of plants and assessing the level of urban greenery,which is common to research.Nowadays,the terms commonly used in the industry to represent urban vegetation planting,such as regional green space area,regional green space rate,regional green coverage rate,regional per capita green space,etc.,consider the situation of urban vegetation from a 2D dimension.Both urban and vegetation exist in 3D,and 3D indicators reflect greenery more scientifically.There are many quantitative studies about natural plants,but the definition of 3D plant quantity is not uniform,the scientific degree of the method and the accuracy of the results are mixed,and the standard of green space evaluation indicators in three-dimensional space is not uniform,so that various 3D plant quantities have not been applied on the ground,which is the special characteristic of the research conducted on 3D vegetation quantity.In this paper,we have reviewed the various opinions on the study of 3D vegetation quantity and proposed a broad definition of 3D vegetation quantity in order to comprehensively consider the level of vegetation planting in regional green areas.The broad 3D vegetation quantity includes three different perspectives of indicators:3D green volume,leaf area density and panoramic visible green index.This study innovatively introduces algorithms and parameterization methods and proposes a regional 3D vegetation quantity estimation method based on plant L-system algorithm.Based on plant growth data,we write the L-system algorithmic rules of different tree species with reference to the four basic branching methods of plants,automate the growth of 50 L-system algorithm plant models.We construct L-system algorithm vegetation models for the current plant growth stages in the district government square green space in Qingshan Jiangtan Park,Wuhan and 5 L-system algorithm vegetation models for the undergraduate dormitory area in Xiangyang,Huazhong Agricultural University.Subsequently,the L-system algorithm plant voxel model with 0.5m and 0.2m voxel accuracy is automatically generated through model point cloud,and 3D green volume are calculated.Based on the L-system algorithm vegetation leaf cluster point cloud model,the voxel-leaf area density model is constructed automatically by algorithm and the regional leaf area density value is calculated.In the virtual space,the virtual panorama image is derived from the L-system algorithm vegetation model and the panoramic visible green index of the regional green space is calculated.Finally,we present the application of the method through two case studies,estimating the 3D vegetation quantity of the green of the district government square area in Qingshan Jiangtan Park,Wuhan at current plant growth stage and the undergraduate dormitory area of the Xiangyang campus of Huazhong Agricultural University in 5 plant growth stages.The accuracy of the method based on L-system algorithm was verified by the comparison of single plant point cloud scanning and the comparison of real and virtual panoramic visible green index.The following conclusions are drawn:(1)The vegetation model with different species of plants constructed based on L-system algorithm can calculate the three-dimensional green amount of regional green space.The spatial IOU(Intersection over Union)between the L-system algorithm plant point cloud model constructed and the single tree point cloud model scanned in the field is 79%,the average IOU(Intersection over Union)between the virtual panoramic visible green index derived from the regional L-system algorithm vegetation model and the corresponding real panoramic visible green index and calculated is 50.18%.The model and green quantity maintain great similarity.(2)At the present stage,the total 3D vegetation volume in the district government square of Qingshan Jiangtan Park is 48158.41m~3,the unit area 3D vegetation volume is about 1.87m~3/m~2.The maximum regional leaf area density is 10 m~2/m~3,the average leaf area density is 4.26m~2/m~3.And the maximum regional panoramic visible green index is91.28%.The average panoramic visible green index was 46.47%.The total 3D vegetation volume of the undergraduate dormitory area Huazhong Agricultural University in Xiangyang is varied from 6910.48m~3 to 52560.15m~3,during the period of plant growth from 3 years to 20 years.The average vegetation leaf area density in the corresponding area is varied from 1.68 m~2/m~3 to 3.06 m~2/m~3.The regional average panoramic visible green index changes from 17.44%to 33.59%.(3)The 3D vegetation quantity of regional green space is mainly affected by the growth years of plants.The growth rate of leaf area density was the fastest,followed by the change of 3D vegetation volume,and the change of panoramic visible green index is the slowest.The 3D vegetation volume is mainly volume,and its size is greatly affected by tree species.The leaf area density is more focused on density,which was affected by the number of branching times.The vertical area of panoramic visible green index is more important,and the size of panoramic green visibility is greatly affected by the degree of spatial vegetation coverage.