A Gaussian measurement model for local interest point based 6 DOF pose estimation

Abstract

One of the main challenges for service robots during operation lies in the handling of unavoidable uncertainties which originate from model and sensor inaccuracies and which are characteristic for realistic application scenarios. Robustness under real world conditions can only be achieved when the dominant uncertainties are explicitly represented and purposefully managed by the robot's control system. We therefore adopt a probabilistic approach in which perception is regarded as a sequential estimation process and follow a Bayesian filtering methodology. Under these assumptions probabilistic models of the robot's perception systems are key.In this paper we shortly describe a model based object recognition and localization system. However, we do not not focus on the 6D pose estimation procedure itself, but on the method to quantify and compute the uncertainty associated with it. We construct a Gaussian approximation of the resulting pose error using the implicit function theorem. It is then used as a proposal density for importance sampling. Our goal is to sample from the measurement model describing 6D object localization based on local features in a Bayesian filtering context.

Bridging the Domain Gap in Satellite Pose Estimation: A Self-Training Approach Based on Geometrical Constraints

“…Object pose estimation aims at recovering the 3D position and 3D rotation of an object in the camera-centered coordinate system. Traditional approaches [28], [29] rely on local features, suffering from texture-less objects and background clutter. Recently, CNN-based methods have dominated most object pose estimation tasks.…”

Real-time monocular object SLAM

“…The pose could be then retrieved by solving the perspective-n-problem [10] from 2D-to-3D correspondences. This has been the basis of a lot of recent object recognition approaches [5,29,7,6,30]. For example, Collet et al [5] build 3D models for 79 objects and use the training images of the objects to build a set of k-d trees to index their SIFT features and Figure 1: System overview: Every video frame is processed by the SLAM tracking thread to locate the camera, and to determine if a new keyframe is added to the map.…”

“…To include objects in SLAM maps, these must be recognized in the images acquired by the robot by computing a rigid-body 3D transformation. A vast research line has provided solutions to this problem [4,5,6,7], but it has been aside from visual SLAM.…”

Table-top scene analysis using knowledge-supervised MCMC

“…al. [52]. For a pose estimate ψ, which corresponds to a scene graph M, we first determine the set of keypoints that have been matched in the object database.…”