Estimation Of Canopy Gap Fraction Based On Porous Media Theory And Computer Graphics

Forestry resources,as one of the most important natural resources for mankind,have always been a hot topic of research.The canopy,as the first part of the plant to come into contact with the external gaseous environment and light,provides a reference point for plant research.The gap fraction of the plant canopy（Gap fraction,GF）is an important indicator to describe the structure of the canopy,which determines the transmission of light to the lower layers of the canopy and is an important attribute affecting the transmission of light and airflow through the canopy.Existing methods for estimating canopy gap fraction based on hemispherical photographs(GF_img)and optical transmission models(GF_light)ignore the natural morphological geometry of the canopy including branches and leaves and are limited by the angle of observation to obtain only local information.Therefore,inspired by porous media theory and computer graphics,this paper undertakes the following work:（1）Using a computer graphics approach in a cross-disciplinary field,four auxiliary conditions are defined to ensure the accuracy of the extracted central region points of each leaf surface.Then,a density-based clustering algorithm with noise（DBSCAN）is used to cluster the central region points of the leaves and obtain the centroids corresponding to each leaf surface to identify single leaves from the laser point cloud.A cylinder fitting algorithm is also used to achieve branch reconstruction based on different radii set by the ecological properties of the tree,thus generating branch（wood）volumes.（2）The equivalent volume of each leaf is estimated using computer graphics moving least squares and cubic polynomial curve fitting algorithms.The concept of equivalent thickness was defined to describe the degree of curl and droop of the leaf,thus representing the interception of light and airflow from the side of the leaf.The blade point cloud is considered as a whole and rotated in the x,y and z axes to obtain the rotation parameter with the lowest axial variance,discern the blade orientation and generate an optimal hexagonal prism based on that orientation to obtain the equivalent volume of the leaf.（3）Estimation of volume-based canopy gap fraction using porous media theory(GF_vol).Treating the canopy as a multi-phase porous medium,the Alphashape convex packet algorithm is used to calculate the volumes of branches,individual leaves and the canopy respectively.The volume-based canopy gap fraction is then the difference between 1 and the ratio of the total volume of the plant leaves and branches to the volume of the canopy.The results of the experiments in this paper show that the estimated gap fraction of five trees and one woodland plot ranged from 0.96 to 0.99,with values closer to 1 indicating better canopy light transmission.Among them,crepe myrtle had the best light transmission and the lowest leaf area index.