Realistic 3D-simulation of large-scale forest scene based on individual tree detection

Abstract

Reconstructing a realistic and large-scale 3D forest scene has potential applications in visual representations and scientific research. Forest scene with explicitly described branches and leaves can provide a more accurate interpretation of interactions between light and canopy. In this study, a large-scale 3D forest scene reconstruction and simulation method is proposed. A series of individual trees with high level of details are generated using parameters derived from allometric equation, which populates plot leaf area index (LAI) into each individual tree. Based on the airborne laser scanning (ALS) data, the height, crown diameter and position of each individual tree are extracted by watershed segmentation algorithm. Finally, an emulation system based on ray-tracing is developed. It provides the capability to simulate RGB and multi-spectral images. These simulated datasets with “ground truth” can be used as benchmark for a variety of applications in remote sensing, forest investigation and computer graphic.