Histogram of Oriented Normal Vectors for Object Recognition with a Depth Sensor

By Shuai Tang, Xiaoyu Wang, Xutao Lv, Tony X. Han, James M. Keller, Zhihai He, M. Skubic, Shihong Lao. 2012

205 citations recorded by the source corpus.

How researchers see this paper

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

Citations in the corpus

Histogram of Oriented Principal Components for Cross-View Action Recognition (2014, methodology): “Experimental results show that our method is able to achieve significantly better accuracy than current state-of-the-art methods for cross-view action recognition and is robust to action speed variations.”
HOPC: Histogram of Oriented Principal Components of 3D Pointclouds for Action Recognition (2014, methodology): “Recently, Oreifej and Liu [19] extended histogram of oriented 3D normals [27] to 4D by adding”
Action Classification with Locality-Constrained Linear Coding (2014, background): “[14] proposed to use histograms of the normal vectors computed from depth images for object recognition.”
3D Object Recognition in Cluttered Scenes with Local Surface Features: A Survey (2014, background): “[121] represented the normal vector orientation as an ordered pair of azimuthal and zenith angles.”
Aligning 3D models to RGB-D images of cluttered scenes (2015, background): “Modiﬁcations to deformable part models [7] for RGB-D images were proposed in [11, 17, 35, 36].”
Human action recognition using meta learning for RGB and depth information (2014, methodology): “HON4D features are the extension of the " Histogram Of Normals " (HON) features [12] and designed to capture changes in the scene structure. extractor constructs features by quantizing the 4D space using a regular 4D extension of a 2D polygon, namely, a " 600-cell Polyhedron " [13].”
Learning with Side Information through Modality Hallucination (2016, unknown): “Many of these methods do so by introducing new depth representations [19, 29, 32, 38, 36], most recently by adding an additional depth network representation into a convolutional network architecture [15, 14, 35].”
Discriminative Orderlet Mining for Real-Time Recognition of Human-Object Interaction (2014, methodology): “Encouraged by the work of using 3D normal vectors for object recognition [7], [12] proposed the histogram of oriented 4D normal features.”
Learning Rich Features from RGB-D Images for Object Detection and Segmentation (2014, background): “Object detection in RGB-D images [20,22,25,35,38], in contrast, focusses on instances, but the typical output is a bounding box.”
Inferring 3D Object Pose in RGB-D Images (2015, methodology): “[11, 17, 34, 35] propose modification to deformable part models [7] to adapt them to the RGB-D domain.”
Simplex-Based 3D Spatio-temporal Feature Description for Action Recognition (2014, background): “A large number of 3D feature description methods are based on spherical coordinate systems [8, 10, 18, 23, 24, 29].”
Grassmannian Representation of Motion Depth for 3D Human Gesture and Action Recognition (2014, background): “Thus, = 5 S = ( ; ; ; 1) = ( n same 1) if we follow the demonstration of Tang et al. [16].”
Cross-modal adaptation for RGB-D detection (2016, background): “[26], [43], [48], [51] propose extensions to deformable part based models [13] to compute additional features from the depth image, and report performance improvements over just using the RGB image.”
Latent-Class Hough Forests for 3D Object Detection and Pose Estimation (2014, methodology): “This representation has further been extended to incorporate depth gradients [42] and normal vectors [74].”
Indoor Scene Understanding with RGB-D Images: Bottom-up Segmentation, Object Detection and Semantic Segmentation (2015, background): “Our work is more similar to that of [46,48], but we differ in the features that we use, and observe that even a simple model with the right features can outperform more complicated approaches.”
Real time action recognition using histograms of depth gradients and random decision forests (2014, background): “[10] proposed histograms of the normal vectors for object recognition in depth images.”
Super Normal Vector for Activity Recognition Using Depth Sequences (2014, background): “As demonstrated in [16], the surface normal provides most shape and structure information of an object in 3D.”
Ieee Transactions on Pattern Analysis and Machine Inteligence 1 Super Normal Vector for Human Activity Recognition with Depth Cameras (unknown): “As demonstrated in [35], the surface normal provides informative shape and structure cues to describe an object in 3D. HON4D [27] followed this observation to extend the surface normal to the 4D space and quantized them by the regular and discriminative learned polychorons.”