Evaluation and Optimisation of Crowd-Based Collection of Trees From 3d Point Clouds

Abstract

Abstract. The term "Crowdsourcing" goes back to Jeff Howe (Howe, 2006) and represents a neologism of the words "crowd" and "outsourcing". Unlike outsourcing, where companies outsource certain tasks to known third parties, crowdsourcing outsources tasks to unknown workers (crowdworkers) on the Internet. This allows companies to access large numbers of workers who would otherwise not be available. In this paper, we will discuss an approach for the crowd-based collection of trees by means of minimum bounding cylinders from 3D point clouds. We will demonstrate the used web-interface and compare the results with reference data. To improve the quality of the results, we collect the data not only once but multiple times. This enables us to implement a so-called “Wisdom of the Crowd” approach where we can identify automatically outliers and derive integrated cylinders. We will show in this paper that this approach increases significantly the quality of the results.

Efficient and Accurate Tree Detection from 3D Point Clouds through Paid Crowdsourcing

“…Majority voting is a simple but effective technique (Zhang et al, 2016) and has been used in various spatial labelling tasks. For example, cropland identification in remote sensing images (Salk et al, 2016), classification of building footprint data (Hecht et al, 2018), accessibility map generation (Liu et al, 2018) and crowd-based labelling of 3D LiDAR point clouds (Koelle et al, 2020). For localization tasks, such as tree annotation, we cannot perform simple majority voting, but compute an average position from the crowdsourced data, e.g., based on a DBSCAN clustering procedure (Walter et al, 2020).…”

“…The annotation of trees from 3D point clouds has already been subject of our previous work (Walter et al, 2020). So far, however, we have only applied our method to a few exemplary, manually selected datasets to demonstrate the feasibility in principle.…”

Individual Tree Detection in Urban ALS Point Clouds with 3D Convolutional Networks

“…The basis for our labeling campaign was a modified version of the tool by [111], by which 3D cylinders can be tightly fit into the point cloud around all points comprising each tree. Since the dataset was too large to work efficiently in the labeling tool, we divided the whole dataset into 100 individual jobs, each 200 × 200 m wide and extending 10 m into their neighbors, thereby creating 20 m overlap regions.…”

The Hessigheim 3D (H3D) Benchmark on Semantic Segmentation of High-Resolution 3D Point Clouds and Textured Meshes from UAV LiDAR and Multi-View-Stereo

“…Whereas 2D imagery can be very well interpreted by non-experts (i.e., crowdworkers), labeling 3D data is much more demanding. Although first investigations were conducted on employing crowdworkers for 3D data annotation (Dai et al, 2017;Herfort et al, 2018;Walter et al, 2020;Kölle et al, 2020), these approaches typically try to avoid deriving a full pointwise annotation. This is achieved either by working on object level or by focusing only on necessary points by exploiting active learning techniques.…”