Selection of Large-Scale 3D Point Cloud Data Using Gesture Recognition

Abstract

Part 7: Perception and Signal ProcessingInternational audienceAn essential task when visualizing and analyzing large-scale 3D point cloud data is the selection of subsets of that data. This presents two challenges, the need for a selection method that is independent of the size of the dataset and how to interact with a 3D space effectively in a digital world still rooted in 2D interaction and visualization. We present an interface for defining volumes to select 3D point cloud data that uses hand gesture control with a Leap Motion device. The use of volumes is scalable to very large datasets, and the use of the Leap Motion gives the user access to the third dimension, facilitating interaction with the point cloud data. We illustrate with a large astronomical data archive hosted on the cloud that is retrieved on as-needed basis

Design space for spatio-data coordination: Tangible interaction devices for immersive information visualisation

“…Currently, there is a range of modalities available to perform interactions in 3D spaces, loosely classifiable into: mouse-based; surface+touch-based (e.g., [3,18,42]); midair-based (e.g., [4,31]); and tangible user interfaces. Each technology is briefly presented in the following in the context of visualisation.…”

CAST: Effective and Efficient User Interaction for Context-Aware Selection in 3D Particle Clouds

“…a sphere with two-handed input, while Bacim et al [5] use the concept of half spaces controlled through 3D gestural input. Similarly, Ren et al [41] and Burgess et al [9] use freehand gestures to control the selection. Dedicated techniques exist also for true volumetric devices [23] and special input surfaces [6].…”

Survey of Object-Based Data Reduction Techniques in Observational Astronomy

“…They simply create new data context consisting of selected data points which then may be effectively visualized. Such selection can be done manually [6], using cubes or other geometric structures [39] or based on distance from the viewpoint. More detailed review of methods dealing with large astronomical datasets only for the purpose of visualization can be found in [18].…”