Recognizing Objects in Range Data Using Regional Point Descriptors

By Andrea Frome, Daniel F. Huber, R. Kolluri, Thomas Bülow, Jitendra Malik. 2004

985 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

DIMENSIONALITY BASED SCALE SELECTION IN 3D LIDAR POINT CLOUDS (2012, background): “More advanced features based on harmonics or spin images (Frome et al., 2004; Golovinskiy et al., 2009) are not necessary since the segmentation task, which indeed requires contextual knowledge, is not tackled in this paper.”
SHOT: Unique signatures of histograms for surface and texture description (2014, methodology): “This is preferable to the solution proposed in [17], i.”
Continuous 3D scan-matching with a spinning 2D laser (2009, methodology): “Each error distribution is computed by running a sliding window of a specified length along the trajectories, aligning the trajectories at the beginning of the window and measuring the error at the end of the window.”
Sliding Shapes for 3D Object Detection in Depth Images (2014, positive): “Keypoint matching: Just as SIFT keypoints for image matching [50], a popular type of algorithms [51–60] is to detect keypoints on a 3D point cloud or a mesh, generate descriptors for the keypoints (e.”
Indoor Scene Understanding with RGB-D Images: Bottom-up Segmentation, Object Detection and Semantic Segmentation (2015, background): “These features are relatively crude and can be replaced by richer features such as spin images [26] or 3D shape contexts [17].”
A Comprehensive Performance Evaluation of 3D Local Feature Descriptors (2015, methodology): “One major limitation of 3DSC is that d sc_a descriptors have to be calculated at each model keypoint due to its ambiguity along the azimuth dimension (Frome et al. 2004; Tombari et al. 2010b).”
VoxNet: A 3D Convolutional Neural Network for real-time object recognition (2015, methodology): “Most of this work uses a pipeline combining various hand-crafted features or descriptors with a machine learning classifier ([10], [11], [12], [13]).”
Tutorial: Point Cloud Library: Three-Dimensional Object Recognition and 6 DOF Pose Estimation (2012, background): “Radius-based surface descriptor (RSD) describes the geometry of points in a local neighborhood by estimating their radial relationships.”
3D free-form object recognition in range images using local surface patches (2004, background): “…1999), the surface point signature (Yamany and Farag, 1999), the harmonic shape image (Zhang and Hebert, 1999), the spherical spin image (Correa and Shapiro , 2001), the 3D point's ''fingerprint'' (Sun and Abidi, 2001) and the 3D shape contexts and harmonic shape contexts (Frome et al., 2004).”
Unsupervised Feature Learning for RGB-D Image Classification (2014, background): “However, almost all the previous work [2, 4, 5, 8] focus on handcrafted feature-based image representation for RGB-D image representation, such as 3-dimensional (3D), Local Binary Patterns (3D-LBP) and .”
3DMatch: Learning Local Geometric Descriptors from RGB-D Reconstructions (2016, methodology): “Many geometric descriptors have been proposed including Spin Images [19], Geometry Histograms [12], and Signatures of Histograms [34], Feature Histograms [29].”
Behavior recognition via sparse spatio-temporal features (2005, background): “Detection of objects in 3D spatial volumes is a distinct problem, see for example [8].”
Rotational Projection Statistics for 3D Local Surface Description and Object Recognition (2013, methodology): “For example, Frome et al. (2004) performed 3D object recognition using the sum of the distances between the scene features (i.e. 3DSC) and their corresponding model features.”
Unique Signatures of Histograms for Local Surface Description (2010, methodology): “As for methods relying on the definition of a full local RF, 3D Shape Context [12] modifies the basic idea of Spin Images by accumulating 3D histograms of points within a sphere centered at the”
FAB-MAP: Probabilistic Localization and Mapping in the Space of Appearance (2008, methodology): “For example, laser or sonar data could be represented using quantized versions of the features proposed in Johnson (1997) Frome et al. (2004) Mozos et al. (2005).”