Evaluating the robustness of point clouds from small format aerial photography over a Pinus radiata plantation

ABSTRACT

Extensive research and operational trials over the past 20 years have led to the operational implementation of airborne laser scanning (ALS)-based forest inventory becoming increasingly common. More recently, digital aerial photography (AP), processed using Structure from Motion Multiview Stereopsis (SfM-MVS) photogrammetry, is emerging as an alternative to ALS. Aerial photography may provide some advantages compared with ALS, including lower deployment and data collection costs, easier access to a variety of platforms and sensors, and the opportunity for forest managers to capture and process the data in-house. This study presents an analysis of point-cloud data derived from airborne small format digital AP, and a comparison with ALS data for a Pinus radiata plantation located in north-east Tasmania. The AP was processed using commercially available photogrammetric software and three different processing strategies. The influence of processing strategy, terrain slope, canopy occlusion, canopy cover, photo-overlap and camera location are investigated in order to characterise the point clouds generated using these methods, and to assess the robustness of the photogrammetric solution to these variables. Our analysis provides strong evidence of the robustness of small format AP-based point clouds in this type of forest: characteristics of the dense point cloud are shown to be largely robust to different photogrammetric processing strategies. Observations regarding the influence of terrain slope, photo-overlap, canopy occlusions, canopy cover and camera location can be used to optimise flight planning and photo-acquisition.

KEYWORDS:

• pine plantations
• forest inventory
• airborne LiDAR
• aerial photography
• 3D point cloud
• structure from motion

Acknowledgements

This research was partially funded through Forest and Wood Products Australia (FWPA) Limited, Research Project PNC326-1314: Deployment and integration of cost-effective, high spatial resolution, remotely sensed data for the Australian forestry industry. Timberlands Pacific Pty Ltd provided site access and field staff support.

Conflicts of interest

The authors declare no conflict of interest.