Jenny Magaly Morocho Toaza
Department of Forest Engineering, Forest Management Planning and Terrestrial Measurements, Faculty of Silviculture and Forest Engineering, Transilvania University of Brașov
1. INTRODUCTION
In recent decades, the need to improve accuracy and efficiency in forest inventories has driven the adoption of new data acquisition technologies. The conventional methods, based on field measurements (like diameter at breast height and total tree height), are still being used, but limitations exist with respect to time, cost, and their ability to account for three-dimensional structural complexity of the forests [1,2].
LiDAR (Light Detection and Ranging) technology has made it possible to create high-density point clouds, allowing for a more detailed look at forest structure [3,4], which has significantly expanded its application in forest inventories [5,6]. In particular, mobile laser scanning (MLS), based on simultaneous localization and mapping (SLAM) algorithms, has emerged as an efficient alternative to terrestrial laser scanning (TLS), offering more flexibility and cutting down field acquisition time [7-9]. Despite these benefits, the use of MLS in multi-temporal forest inventories faces a major challenge due to seasonal phenological variability [10-13]. The differences between leaf-on and leaf-off conditions generate significant changes in the observable structure of the forest, affecting laser penetration and the distribution of returns [10,13]. During the leaf-on period, the presence of leaves and undergrowth vegetation increases occlusion and limits the visibility of structural elements such as trunks, while in leaf-off conditions, the detection of woody structures and the ground is favored [12,14]. These differences have a direct impact on point density and the geometric representation of the forest, which can lead to biases in the analysis if not properly addressed [10]. […]
