Volumetric 3D mapping in real-time on a CPU

By Frank Steinbrücker, Jürgen Sturm, Daniel Cremers. 2014

174 citations recorded by the source corpus.

Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Aerospace Engineering

In Computer Vision and Pattern Recognition

Citations in the corpus

Deep Level Sets: Implicit Surface Representations for 3D Shape Inference (2019, methodology): “models (GANs) using long short-term memory (LSTM) for reconstructing voxels or point clouds achieving state-of-theart results. [28] proposed the use of conditional GANs in an unsupervised setting and [31] proposed the use of octrees. An important drawback of GAN-based methods is that they are comput”
Hierarchical Surface Prediction for 3D Object Reconstruction (2017, methodology): “It is inspired from octree formulations [5], [31] used in traditional multi-view reconstruction approaches.”
Octree Generating Networks: Efficient Convolutional Architectures for High-resolution 3D Outputs (2017, background): “Octrees have a long history in classical 3D reconstruction and depth map fusion [21, 1, 11, 34, 6, 31].”
ElasticFusion: Real-time dense SLAM and light source estimation (2016, methodology): “To evaluate the trajectory estimation performance of our approach we test our system on the RGB-D benchmark of Sturm et al. (2012). This benchmark provides synchronised ground truth poses for an RGB-D sensor moved through a scene, captured with a highly precise motion capture system.”
VolumeDeform: Real-Time Volumetric Non-rigid Reconstruction (2016, methodology): “al time. Follow-up work extended the underlying data structures and depth fusion algorithms in order to provide better scalability for handling larger scenes [36] and a higher reconstruction quality [7,8]. While these approaches achieve impressive results on static environments, they do not reconstr”
ElasticFusion: Dense SLAM Without A Pose Graph (2015, result): “This is in contrast to other dense reconstruction systems which don’t strictly perform both tracking and mapping in real-time [18, 19].”
ReFusion: 3D Reconstruction in Dynamic Environments for RGB-D Cameras Exploiting Residuals (2019, methodology): “[16] use an Octree instead of a voxel grid.”
Implicit Surface Representations As Layers in Neural Networks (2019, background): “[29] proposed the use of conditional GANs in an unsupervised setting and [32] proposed the use of octrees.”
Voxblox: Incremental 3D Euclidean Signed Distance Fields for on-board MAV planning (2016, methodology): “All state estimation, reconstruction, planning, and control runs entirely on-board on the Intel i7 2.1 GHz CPU without any reliance of any external equipment or infrastructure.”
BOP: Benchmark for 6D Object Pose Estimation (2018, methodology): “The 3D object models were created using KinectFusion-like systems for 3D surface reconstruction [26,33].”
Very High Frame Rate Volumetric Integration of Depth Images on Mobile Devices (2015, methodology): “This important issue has been addressed in recent years by sparse volumetric representations like [28, 2, 24] and [16].”
T-LESS: An RGB-D Dataset for 6D Pose Estimation of Texture-Less Objects (2017, methodology): “The reconstructed models were created using fastfusion, a volumetric 3D mapping system by Steinbr¨ucker et al. [44].”
Real-Time Large-Scale Dense 3D Reconstruction with Loop Closure (2016, methodology): “Their memory requirements have been reduced by an order of magnitude more recently [13–17], rendering them attractive for large-scale mapping.”