Dense visual SLAM for RGB-D cameras

Abstract

Abstract-In this paper, we propose a dense visual SLAM method for RGB-D cameras that minimizes both the photometric and the depth error over all pixels. In contrast to sparse, feature-based methods, this allows us to better exploit the available information in the image data which leads to higher pose accuracy. Furthermore, we propose an entropy-based similarity measure for keyframe selection and loop closure detection. From all successful matches, we build up a graph that we optimize using the g2o framework. We evaluated our approach extensively on publicly available benchmark datasets, and found that it performs well in scenes with low texture as well as low structure. In direct comparison to several stateof-the-art methods, our approach yields a significantly lower trajectory error. We release our software as open-source.

LSD-SLAM: Large-Scale Direct Monocular SLAM

“…While direct image alignment is well-established for RGB-D or stereo sensors [14,4], only recently monocular direct VO algorithms have been proposed: In [24,20,21], accurate and fully dense depth maps are computed using a vari- Fig. 2: In addition to accurate, semi-dense 3D reconstructions, LSD-SLAM also estimates the associated uncertainty.…”

“…In [14], a pose graph based RGB-D SLAM method is proposed, which also incorporates geometric error to allow tracking through scenes with little texture. To account for scale-drift arising in monocular SLAM, [23] proposed a keypointbased monocular SLAM system which represents camera poses as 3D similarity transforms instead of rigid body movements.…”

“…from occlusions or reflections, different weighting-schemes [14] have been proposed, resulting in an iteratively reweighted least-squares problem: In each iteration, a weight matrix W = W(ξ (n) ) is computed which down-weights large residuals. The iteratively solved error function then becomes…”

“…where the photometric residual r 2 p and σ 2 rp is defined as in (13) - (14). The depth residual and its variance is computed as…”

Vision based persistent localization of a humanoid robot for locomotion tasks

“…Here, a photo-consistency error was measured between all pixels of consecutive images in order to compute the a-posteriori likelihood of the camera motion. The idea was extended with visual SLAM capabilities by Kerl et al (2013a), who added a depth error to the cost function in order to achieve scene reconstruction and loop closure. More recently, Endres et al (2014) approached a geometric solution including robust matching of visual features using the sensor input as landmark positions in order to compute the 3D-to-3D relations for camera motion estimation.…”