Development of a Probabilistic Sensor Model for a 3D Imaging System

Abstract

The use of high frame-rate 3D imaging systems for construction equipment control is part of a broader research effort at the National Institute of Standards and Technology investigating innovative technologies for automated construction, performance and evaluation of 3D imaging systems, and construction object recognition and tracking. A probabilistic measurement model is desired for these imaging systems to better understand and describe the instrument performance, and to create synthetic data sets for algorithm development, training and testing. The initial development of a probabilistic sensor model and 3D image simulator based on an open-source graphics package is presented in this paper. KEYWORDS3D Imaging System, Flash LADAR, 3D Range Camera, Sensor Model.

Tracking and Classifying Objects on a Conveyor Belt Using Time-of-Flight Camera

“…Theobalt et al (2007) have proposed that super resolution methods typically used for color images produce high resolution 3D data of a scene using 3D range camera. To effectively reduce noise in the 3D spatial data acquired by range cameras, a probabilistic sensor model was proposed by Lytle (2009). The work shows, that a point cloud of a synthetic 3D image generated by the probabilistic model appears representative of the dataset that would be captured by an actual 3D range instrument.…”

3D Imaging System for the Intelligent Excavation System (IES)

“…The technology for the effective modeling of objects around the work environment and for automatic recognition is the first indispensable step in an automation or intelligent process [2]. It started with simple sensors and produced many results, including a vision system using a camera and a 3D modeling system using a broad-area 3D laser scanner, continuing efforts to realize better environment recognition.…”