On the Repeatability and Quality of Keypoints for Local Feature-based 3D Object Retrieval from Cluttered Scenes

By Ajmal Saeed Mian, Mohammed Bennamoun, Robyn A. Owens. 2009

452 citations recorded by the source corpus.

Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition, Information Retrieval & Search, 3D Vision & Pose

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

In Information Retrieval & Search

In 3D Vision & Pose

Citations in the corpus

VoxelNet: End-to-End Learning for Point Cloud Based 3D Object Detection (2017, methodology): “Some of the earlier methods for feature representation are [41, 8, 7, 19, 42, 35, 6, 27, 1, 36, 2, 25, 26].”
Histogram of Oriented Principal Components for Cross-View Action Recognition (2014, background): “(1) Constant spatial scale for all subjects, (2) ratio of subject’s height as the spatial scale, and (3) automatic spatial scale selection[59].”
HOPC: Histogram of Oriented Principal Components of 3D Pointclouds for Action Recognition (2014, background): “[17] to determine the optimal spatial scale for each keypoint.”
Taskonomy: Disentangling Task Transfer Learning (2018, background): “They are often invariant to informative (but possibly distracting) features such as textures [50, 39, 25, 47, 16].”
SUN RGB-D: A RGB-D scene understanding benchmark suite (2015, unknown): “There are many existing RGB-D datasets available [54] [47] [1] [25] [44] [60] [49] [45] [29] [66] [57] [52] [46] [16] [21] [73] [35] [67] [3] [41] [10] [63] [42] [64] [65] [48] [12] [33] [8] [50] [26] [6].”
SHOT: Unique signatures of histograms for surface and texture description (2014, methodology): “This task is accomplished by 3D detectors [5, 6, 7], whose aim is to single out points that are distinctive, to allow for effective description and matching, as well as repeatable with respect to vantage point variations and noise.”
A combined texture-shape descriptor for enhanced 3D feature matching (2011, background): “Several approaches have been proposed for 3D feature point detection and description [1–7].”
3D Object Recognition in Cluttered Scenes with Local Surface Features: A Survey (2014, methodology): “technique include [4], [9], [44], [46], [70] and [127].”
Rotational Projection Statistics for 3D Local Surface Description and Object Recognition (2013, methodology): “The recognition performance of the adaptive-scale RoPS based algorithm was better than that re-ported in (Mian et al., 2010), which means that our RoPS descriptor was more descriptive than the descriptor used in (Mian et al., 2010).”
A Comprehensive Performance Evaluation of 3D Local Feature Descriptors (2015, background): “mation content are first identified and their associated scales (spatial extents) are determined (Mian et al. 2010; Tombari et al. 2013).”
Model globally, match locally: Efficient and robust 3D object recognition (2010, background): “[10, 11], which consists of 50 scenes taken with a Minolta range scanner.”
Harris 3D: a robust extension of the Harris operator for interest point detection on 3D meshes (2011, background): “[18] related the repeatability of keypoints (extracted from partial views of an object) with a quality measure based upon principal curvatures.”
Image Matching from Handcrafted to Deep Features: A Survey (2020, methodology): “Mian et al. (2010) proposed an automatic scale selection technique for extracting scale invariant features.”
SpinNet: Learning a General Surface Descriptor for 3D Point Cloud Registration (2020, methodology): “Given a specific point p ∈ P in a local surface, we first estimate a reference axis np oriented to the viewpoint [40, 1] from its neighbouring point set P = {pi : ‖pi − p‖ 2 ≤ R} within a support radius R.”
Voting-based pose estimation for robotic assembly using a 3D sensor (2012, background): “Several feature descriptors and matching algorithms have been proposed [12], [13], [14], [15], [16].”
Recognizing Objects in 3D Point Clouds with Multi-Scale Local Features (2014, methodology): “The UWA dataset is currently regarded as the most popular benchmark for 3D object recognition [4,10,11,13,19,27,32].”
Learning to Detect Good 3D Keypoints (2017, methodology): “…detectors, we evaluate those available in PCL, namely ISS, Harris3D and NARF (Steder et al. 2011), alongside with the algorithm proposed by Mian et al. (2010) and referred to as KPQ in Tombari et al. (2013), which, together with ISS, turned out particularly effective in the experiments…”