Stand density estimators based on individual tree detection and stochastic geometry

Abstract

Individual tree detection methods leave smaller trees hiding below larger ones undetected. This is a problem for remote sensing forest inventories, leading e.g. to severe underestimation of stand density. We develop new methods of formulating the probability of detecting individual trees-the detectability-based on stochastic geometry, and use them to derive estimators of stand density. We assume that a tree remains undetected if the center point of the crown falls within an erosion set based on the larger tree crowns. These estimators allow the tree to be undetected even if a portion of its crown would be visible, taking into account possible differences in accuracy of remote sensing data and detection algorithms. The behaviour of these estimators is quantified using 36 field plots, and compared to a previously proposed estimator. The best estimator according to the data used, allowing trees to be undetected when 40 percent or more of crown radius is hidden, performs well compared to the estimator formed directly from the number of algorithmically detected trees. It produces a 54 percent reduction in RMSE and shifts the mean of errors notably towards zero in the modelling data. Small variations in allowed visible crown radius do not seem to impact the accuracy of the estimates. Generalization of the results remains as a topic of future research.

Extraction of Sample Plot Parameters from 3D Point Cloud Reconstruction Based on Combined RTK and CCD Continuous Photography

“…In addition to the location and size (DBH and tree height) of the trees, there are many factors that can objectively reflect stand features such as the tree density, average diameter at breast height, average height of the stand, and stand spatial structure parameters (angle scale, size ratio, and mixed degree) [40].…”

A Comparison of Two Tree Detection Methods for Estimation of Forest Stand and Ecological Variables from Airborne LiDAR Data in Central European Forests

“…Regardless of the detection techniques, however, stand variables that are directly linked to the number of detected trees are often underestimated due to problems in detecting suppressed and understory trees (e.g., growing stock). To overcome this problem, semi-ITD approaches and several techniques for modeling understory trees have been proposed [19][20][21][22][23].…”

Delineating forest stands from grid data

“…The fact that small trees are often hided by the crowns of large trees has decreased the accuracy of forest inventories based on individual tree detection. However, new methods and algorithms are being developed which mitigate this problem (e.g., Kansanen et al 2016).…”