WK 49 - An Open Source Workflow to Quantify Canopy Structure from Aerial Imagery with OpenDroneMap

Session Description: Unoccupied aerial systems (UAS) technology is revolutionizing plant ecology by providing high-resolution imagery across large spatial extents. Achieving the full potential of UAS for ecology will require extracting quantitative data from imagery. Structural data, including canopy height, volume, and size measurements of individual plants, are a type of remotely sensed data with broad relevance for ecosystem, population, and community ecology. Advances in computer vision have enabled structural measurements from aerial images, namely structure-for-motion (SfM). SfM uses overlapping images, taken from different angles, to construct three-dimensional point clouds, which can then be used to develop structural information. In ecosystems with open canopies, SfM point clouds have comparable accuracy to point clouds developed using aerial lidar. However, SfM point clouds can be created with RGB cameras mounted on drones that are orders of magnitude cheaper than UAS-borne lidar systems. This relatively low cost will render these technological advances accessible to a wide range of ecologists.

In this workshop, we will present an open source workflow to extract structural information from RGB cameras, using OpenDroneMap. OpenDroneMap is a freely available processing tool with a large and growing user community. This open source tool has potential for broader use by ecologists. We will demonstrate how OpenDroneMap can create structural products, using drone imagery acquired with a relatively inexpensive RGB camera. Our demonstration will also provide R code to convert point clouds into ecologically-relevant information, including segmented crowns of individual plants and canopy height models. Session participants will also have the opportunity to process their own data.

Our workshop will increase the utility of commonly-collected drone imagery by providing a low-cost, open source workflow to model ecosystem structure. Resultant products will be useful for many ecological applications, from measuring habitat quality to tracking demographic changes in plant populations.