Performance Evaluation of 3D Keypoint Detectors
Citations
667 citations
Cites background from "Performance Evaluation of 3D Keypoi..."
...Some pipelines use point-to-point matches based on local geometric descriptors [16, 40], others define correspondences on pairs or tuples of points [1, 8, 26, 29]....
[...]
602 citations
Additional excerpts
...A performance evaluation of 3D keypoint detection algorithms has been recently proposed in [12]....
[...]
563 citations
503 citations
474 citations
Cites background from "Performance Evaluation of 3D Keypoi..."
...Readers are referred to Tombari et al. (2013) for further discussion on other adaptive-scale detectors....
[...]
...D case, Tombari et al. (2013) presented a thorough evaluation of several state-of-the-art 3-...
[...]
...D keypoint detectors, Tombari et al. (2013) provided an excellent survey on the state-of-the-art methods and a detailed evaluation of their performances....
[...]
References
46,906 citations
14,708 citations
"Performance Evaluation of 3D Keypoi..." refers background or methods in this paper
...Likewise (Unnikrishnan and Hebert 2008), also MeshDoG (Zaharescu et al. 2009) deploys the 3D mesh as the representation adopted to build the scale-space, which, in turn, is created by applying different normalized Gaussian derivatives through the Difference-of-Gaussians (DoG) operator, a well-known approximation of the normalized Laplacian (Lowe 2004)....
[...]
...proposed in Lowe (2004), non-corner responses are pruned...
[...]
...…et al. 2009) deploys the 3D mesh as the representation adopted to build the scale-space, which, in turn, is created by applying different normalized Gaussian derivatives through the Difference-of-Gaussians (DoG) operator, a well-known approximation of the normalized Laplacian (Lowe 2004)....
[...]
...2009) deploys the 3D mesh as the representation adopted to build the scale-space, which, in turn, is created by applying different normalized Gaussian derivatives through the Difference-of-Gaussians (DoG) operator, a well-known approximation of the normalized Laplacian (Lowe 2004)....
[...]
...6http://www.caam.rice.edu/software/ARPACK/. to solve the sparse generalized eigenvalues problem and UMFPACK7 to perform the LU factorization of the mesh Laplacian operator L....
[...]
3,359 citations
"Performance Evaluation of 3D Keypoi..." refers background or methods in this paper
...2010; Chen and Bhanu 2007; Zhong 2009; Novatnack and Nishino 2008; Unnikrishnan and Hebert 2008; Akagunduz and Ulusoy 2007; Zaharescu et al. 2009; Fadaifard and Wolberg 2011; Knopp et al. 2010; Castellani et al. 2008; Sun et al. 2009). Compared to its 2D counterpart, and to 3D descriptors as well, research on 3D detectors has started much more recently (mainly in the past 3 or 4 years) and, to date, only limited taxonomic and evaluation work has been carried out, with experimental comparison usually performed separately within (and relatively to) each specific proposal. The only relevant works in this respect are described in Bronstein et al. (2010), Boyer et al. (2011), which propose an experimental evaluation of 3D detectors and descriptors focused on the 3D shape retrieval scenario, and a preliminary version of this paper, which was presented in Salti et al....
[...]
...2000; Mikolajczyk et al. 2005) and mainly focused on the 3D object recognition scenario, which is peculiarly characterized by the presence of occlusions and clutter. Such a scenario differs from that addressed by Bronstein et al. (2010), Boyer et al. (2011), as 3D shape retrieval is not required to deal with occlusion, clutter and viewpoint changes, large intraclass shape variations being instead the main nuisance to be dealt with....
[...]
...For this reason, and as done in Mikolajczyk et al. (2005), we chose to use the default parameters supplied by the authors rather than tuning them....
[...]
...As discussed in Mikolajczyk et al. (2005), these differences may have an undesired impact on the repeatability scores: if the number of keypoints is large, many of them may be considered repeatable by accident and not because of the design of the detector....
[...]
...2010; Chen and Bhanu 2007; Zhong 2009; Novatnack and Nishino 2008; Unnikrishnan and Hebert 2008; Akagunduz and Ulusoy 2007; Zaharescu et al. 2009; Fadaifard and Wolberg 2011; Knopp et al. 2010; Castellani et al. 2008; Sun et al. 2009). Compared to its 2D counterpart, and to 3D descriptors as well, research on 3D detectors has started much more recently (mainly in the past 3 or 4 years) and, to date, only limited taxonomic and evaluation work has been carried out, with experimental comparison usually performed separately within (and relatively to) each specific proposal. The only relevant works in this respect are described in Bronstein et al. (2010), Boyer et al. (2011), which propose an experimental evaluation of 3D detectors and descriptors focused on the 3D shape retrieval scenario, and a preliminary version of this paper, which was presented in Salti et al. (2011). A parallel line of research has targeted the evaluation of 2D detectors and descriptors on 3D objects (Moreels and Perona 2007)....
[...]
2,942 citations
"Performance Evaluation of 3D Keypoi..." refers background in this paper
...4, the common structure of adaptivescale detectors includes building a scale-space defined on the surface, thus directly extending to the case of 3D data the well-known concept defined for 2D images (Lindeberg 1998)....
[...]
2,798 citations