#### DMCA

## Gaussian sphere Gaussian

### Citations

2447 | A combined corner and edge detector
- Harris, Stephens
- 1988
(Show Context)
Citation Context ...um points of the HKS are considered as feature points. Other methods are extracting feature points locally. Sipiran and Bustos [42] proposed an extension of the seminal Harris corner detection scheme =-=[16]-=- which considers a quadratic patch as the vicinity of a vertex to extract feature points. Litman et al. [25] represent the surface as a tree using a graph-based diffusion formulation and then compute ... |

1519 |
Multidimensional binary search trees used for associative searching
- Bentley
- 1975
(Show Context)
Citation Context ...es and tessellated into 1996 cells, STEP 2 For a given model, compute the surface normal at each vertex map the normals on the surface of the Gaussian accumulator, STEP 3 Using the k-d tree algorithm =-=[3]-=-, find the corresponding cell and increment the count for this cell of the accumulator, STEP 4 Sort the voted Gaussian accumulator in order of decreasing count value (see figure 12 for examples of suc... |

573 |
Gool, “SURF: Speeded up robust features
- Bay, Tuytelaars, et al.
- 2006
(Show Context)
Citation Context ...ussian Sphere and the Gaussian Accumulator 5 as spatial information. A method which uses a visual dictionary and spatial knowledge is presented in [24]. After voxelizing the 3D surface, SURF features =-=[2]-=- are extracted from each voxel, used to build a visual vocabulary, and then a Hough Transform-like voting method is used to build the histogram for each set of words. An extension of this work is prop... |

85 | Shape Google: geometric words and expressions for invariant shape retrieval
- Bronstein, Bronstein, et al.
(Show Context)
Citation Context ... built for each model. These distributions are then compared using the Kullback-Leiber divergence measure. Recently, in addition to global descriptors, local descriptors have also attracted attention =-=[7,5,29]-=-. In such approaches, a mesh is described by a set of local descriptors, each of them characterizing a specific neighborhood of vertices in the mesh. Each mesh can be described by a different number o... |

68 | 3D object recognition from range images using local feature histograms
- Hetzel, Leibe, et al.
(Show Context)
Citation Context ...o recognize the objects. A robust recognition rate is obtained by using Kullback-Leibler and likelihood matching approaches. The method is still sensitive to noise. The global descriptor presented in =-=[17]-=- calculates the surface normal and the curvature value. It then takes locally the depth value of each pixel of the object and groups them in a multidimensional histogram as a descriptor. Another globa... |

49 | Sparse points matching by combining 3d mesh saliency with statistical descriptors
- Castellani, Cristani, et al.
(Show Context)
Citation Context ...cted as the interest points. It is in fact a Mesh-DoG operator. A Mesh-SIFT operator was proposed in [27] for which local features are scale-space extrema. Another method based on DoG was proposed in =-=[6]-=-. They first apply Gaussian filters to the mesh and find differences between the octaves to obtain DoG function. This procedure is then repeated on a different downsampled version of the input surface... |

26 |
Dense sampling and fast encoding for 3D model retrieval using bagof-visual features,
- Furuya, Ohbuchi
- 2009
(Show Context)
Citation Context ... of depth images captured from different views of 3D objects. These descriptors are then quantized to prepare the dictionary words and the BoW approach is used for shape retrieval. Furuya and Ohbuchi =-=[11]-=- extended this approach to extract more local visual features using a dense depth image sampling. When spatial information is taken into account in the BoW approach, the result can be more accurate, s... |

22 | Matching shapes by eigendecomposition of the Laplace-Beltrami operator
- Dubrovina, Kimmel
(Show Context)
Citation Context ... find the area of a triangle, we simply identify two adjacent side vectors of the triangle and then the area is half the cross product norm of these two adjacent vectors [28]: areaj = 1 2 juj vj j, =-=(8)-=- where, uj and vj are the two adjacent side vectors of triangle j.The surface normal at i is mapped on the Gaussian accumulator and the cell corresponding to ni is incremented by Si (instead of 1). Th... |

11 | Scale invariant features for 3dmesh models
- Darom, Keller
- 2010
(Show Context)
Citation Context ...re and the Gaussian Accumulator 3 structure of even the simple 3D meshes such as spheres, and cylinders. Such feature sets are called shape descriptors which can describe shapes locally or globally =-=[7]-=-. A global descriptor represents a whole mesh as a feature vector which includes some information such as area, volume, statistical moments, Fourier coefficients, eigendecomposition of Laplac... |

7 | A bag of features approach for 3d shape retrieval
- Fehr, Burkhardt
- 2009
(Show Context)
Citation Context ...pth SIFT, the extension of local SIFT descriptor [26] in 3D meshes. 2.4 Bag of Words Approaches The Bag of Words scheme, which is based on local descriptors, is used to retrieve images [43] or meshes =-=[5, 9]-=- from a large database of models. In a dataset of images or meshes, the BoW method first defines a feature dictionary from local descriptors and then quantizes the features to dictionary words. Afterw... |

6 |
N.M.: Object recognition using 3D SIFT in complex CT volumes
- Flitton, Breckon, et al.
- 2010
(Show Context)
Citation Context ...tion and translation. As 3D data sets are common in medical image processing applications, the 3D-SIFT scheme was proposed in [38]. This method was then applied to object recognition in CT volumes by =-=[10]-=-. In [7], two scale and rotation invariant local descriptors in 3D meshes are presented: 1) the Scale Invariant Spin Images, a scale invariant extension of original Spin Images [20]; 2) the Local Dept... |

6 | Rotational projection statistics for 3d local surface description and object recognition
- Guo, Sohel, et al.
(Show Context)
Citation Context ... a surface with a reduced amount of data. A recent comprehensive evaluation of 3D local descriptors as well as a survey on local surface features have been proposed in [14] and [13], respectively. In =-=[15]-=-, a new local reference frame has also been introduced in order to extract local features for 3D recognition. In the following we present two wellknown local descriptors. The Spin Image mesh descripto... |

4 |
3D object recognition in cluttered scenes with local surface features: A survey
- Guo, Bennamoun, et al.
- 2014
(Show Context)
Citation Context ...p to describe a mesh or a surface with a reduced amount of data. A recent comprehensive evaluation of 3D local descriptors as well as a survey on local surface features have been proposed in [14] and =-=[13]-=-, respectively. In [15], a new local reference frame has also been introduced in order to extract local features for 3D recognition. In the following we present two wellknown local descriptors. The Sp... |

2 | M.S.: Feature based 3D object recognition using articial neural networks
- Baareh, Sheta, et al.
- 2012
(Show Context)
Citation Context ...s. The most successful methods are based on statistical machine learning theory such as kernel methods. In order to use learning approaches, some proposed methods rely on a feature extraction process =-=[1]-=-. In particular, [31,39] try to extract features from a set of 2D images taken from different views. Then, based on these features, the object is recognized without considering its 3D geometry [1]. Si... |

2 |
N.D.: Method for registration of 3-D shapes. Robotics-DL tentative
- Besl, McKay
- 1992
(Show Context)
Citation Context ...Another global descriptor called Potential Well Space Embedding is proposed in [41]. It is based on comparing a 7D error function using the properties of the ICP (Iterative Closest Point) algorithm =-=[4]-=-. Object recognition is performed by comparing the feature vectors of runtime data with those of the preprocessed database. The Spherical Harmonic Descriptor by Kazhdan et al. [21] is another global d... |

2 |
R.: Shape analysis using the auto diusion function. Computer Graphics Forum 28(5
- Gebal, Bærentzen, et al.
- 1405
(Show Context)
Citation Context ... the eigenfunctions are computed and the feature points are the local extremum points of these functions. Two other methods based on LaplaceBeltrami operator are proposed to find surface features. In =-=[12]-=- and [44], a Heat Kernel Signature (HKS) function is calculated over the surface using the Laplace4 Zahra Toony et al. Beltrami operator and extremum points of the HKS are considered as feature points... |

1 |
N.M.: A comprehensive performance evaluation of 3d local feature descriptors
- Guo, Bennamoun, et al.
- 2015
(Show Context)
Citation Context ...ssing step to describe a mesh or a surface with a reduced amount of data. A recent comprehensive evaluation of 3D local descriptors as well as a survey on local surface features have been proposed in =-=[14]-=- and [13], respectively. In [15], a new local reference frame has also been introduced in order to extract local features for 3D recognition. In the following we present two wellknown local descriptor... |