Real-Time Camera Tracking and 3D Reconstruction Using Signed Distance Functions

Abstract

Abstract-The ability to quickly acquire 3D models is an essential capability needed in many disciplines including robotics, computer vision, geodesy, and architecture. In this paper we present a novel method for real-time camera tracking and 3D reconstruction of static indoor environments using an RGB-D sensor. We show that by representing the geometry with a signed distance function (SDF), the camera pose can be efficiently estimated by directly minimizing the error of the depth images on the SDF. As the SDF contains the distances to the surface for each voxel, the pose optimization can be carried out extremely fast. By iteratively estimating the camera poses and integrating the RGB-D data in the voxel grid, a detailed reconstruction of an indoor environment can be achieved. We present reconstructions of several rooms using a hand-held sensor and from onboard an autonomous quadrocopter. Our extensive evaluation on publicly available benchmark data shows that our approach is more accurate and robust than the iterated closest point algorithm (ICP) used by KinectFusion, and yields often a comparable accuracy at much higher speed to feature-based bundle adjustment methods such as RGB-D SLAM for up to medium-sized scenes.

Scale-aware navigation of a low-cost quadrocopter with a monocular camera

“…Previous work on autonomous quadrocopter flight has explored lightweight laser scanners [5], RGB-D sensors [8,9] or stereo rigs [16] mounted on a quadrocopter as primary sensors. While these sensors provide absolute scale of the environment, their drawback is a limited range and large weight, size, and power consumption when compared to a monocular set-up.…”

“…For solving the SLAM problem on MAVs, different types of sensors such as laser range scanners [5], monocular cameras [6], stereo cameras [7], and RGB-D sensors [8,9] have been explored in the past. In our point of view, monocular cameras have two major advantages Figure 1: A low-cost quadrocopter navigates in unstructured environments using the front camera as its main sensor.…”

Chisel: Real Time Large Scale 3D Reconstruction Onboard a Mobile Device using Spatially Hashed Signed Distance Fields

“…Many state-of-the-art real-time 3D reconstruction algorithms [2,[11][12][13][14] compute a truncated signed distance field (TSDF) [15] of the scene. The TSDF stores a discretized estimate of the distance to the nearest surface in the scene.…”

Tracking objects with point clouds from vision and touch

“…Many variants of the Iterative Closest Point (ICP) algorithm have been developed to locate and track objects in point clouds [17]. In conjunction with ICP, signed distance functions (SDF) have proven a valuable tool for reasoning in a continuous and smooth way about the geometries of objects and scenes, and have proven invaluable in object tracking, and simultaneous localization and mapping (SLAM) [2] [20] [15] [23].…”