The Representation, Recognition, and Locating of 3-D Objects

By Olivier D. Faugeras, Martial Hebert. 1986

How researchers see this paper

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

1. Bo Li
2. Andrew Zisserman
3. Luc Van Gool
4. Xiaogang Wang
5. Martial Hebert
6. Wanli Ouyang
7. Yi Yang
8. Dacheng Tao

Citations in the corpus

• 3D fully convolutional network for vehicle detection in point cloud (2016, methodology): “], [31], sliding window [30], random sampling [13], or the recently popular Region Proposal Network (RPN) [26]. For the classification stage, research have been drawn to features including shape model [6], [13] and geometry statistic features [22], [27], [29], [31]. Sparsing coding [4], [15] and deep”
• Vehicle Detection from 3D Lidar Using Fully Convolutional Network (2016, methodology): “To classify the candidate data, some early researches assume known shape model and match the model to the range scan data [6, 13].”
• Least squares 3D surface and curve matching (2005, methodology): “During that time two other similar methods had been developed independently based on unit quaternions (Horn, 1987; Faugeras and Hebert, 1986), but as pointed out by Horn et al.”
• Using Spin Images for Efficient Object Recognition in Cluttered 3D Scenes (1999, methodology): “The usual method for dealing with clutter is to segment the scene into object and nonobject components [1], [7]; naturally, this is difficult if the position of the object is unknown.”
• 3D free-form object recognition in range images using local surface patches (2004, background): “In the early years of 3D computer vision (Besl and Jain, 1985; Chin and Dyer, 1986), the representation schemes included Wire-Frame, Constructive Solid Geometry (CSG), Extended Gaussian Image (EGI), Generalized Cylinders and planar faces (Bhanu, 1984; Faugeras and Hebert, 1986).”
• Invariant Visual Object and Face Recognition: Neural and Computational Bases, and a Model, VisNet (2012, unknown): “Other schemes use 2D surface patches as their primitives (Dane and Bajcsy, 1982; Brady et al., 1985; Faugeras and Hebert, 1986; Faugeras, 1993).”
• Active Contours (2000, background): “In some cases subsequent processes can tolerate errors - gaps in contours and spurious fragments - and this is particularly true of certain approaches to object recognition, for instance (Ballard, 1981; Grimson and Lozano-Perez, 1984; Faugeras and Hebert, 1986; Mundy and Heller, 1990).”
• A Review of Medical Image Registration (1993, methodology): “Closed-form solutions using the singular value decomposition (SVD) of a matrix [9], the eigenvalue-eigenvector decomposition of a matrix [91], and unit quaternions [51, 90] have been independently rediscovered.”
• 3D Object Modeling and Recognition Using Local Affine-Invariant Image Descriptors and Multi-View Spatial Constraints (2006, unknown): “…a given set of geometric constraints have been proposed in the past, including interpretation tree—or alignment—techniques (Ayache and Faugeras, 1986; Faugeras and Hebert, 1986; Grimson and Lozano-P´erez, 1987; Huttenlocher and Ullman, 1987; Lowe, 1987), geometric hashing (Lamdan and Wolfson,…”
• ICP Registration Using Invariant Features (2002, methodology): “Repeat for k = 1::kmax or until termination criteria is met (a) Build up the set of correspondences C = SNs i=1f(Tk 1(si);CP(Tk 1(si);M))g (b) Compute the new transformation Tk that minimizes mean square error between point pairs in C [12, 15].”
• Learning Recognition and Segmentation Using the Cresceptron (1997, background): “, Faugeras & Hebert 1986 [14], Huttenlocher & Ullman 1987 [27]), constrained search (e.”