Performance Evaluation of 3D Keypoint Detectors

Tombari, Federico; Salti, Samuele; Stefano, Luigi Di

doi:10.1007/s11263-012-0545-4

Cited by 337 publications

(214 citation statements)

References 17 publications

Supporting

Mentioning

211

Contrasting

Unclassified

Order By: Relevance

“…In the case of TLS, the obtained unstructured point clouds (individual scans) are cleaned, edited and aligned in a common local reference system. The alignment process can be automatically carried out through Iterative Closest Point (ICP) points [47], by automatic recognition of objects [48] or using artificial targets [49]. The 3D point cloud obtained with milimetric resolution can include the intensity levels (I) and/or the color information (RGB).…”

Section: Methodsmentioning

confidence: 99%

Mobile LiDAR System: New Possibilities for the Documentation and Dissemination of Large Cultural Heritage Sites

et al. 2017

View full text Add to dashboard Cite

Abstract:Mobile LiDAR System is an emerging technology that combines multiple sensors. Active sensors, together with Inertial and Global Navigation System, are synchronized on a mobile platform to produce an accurate and precise geospatial 3D point cloud. They allow obtaining a large amount of georeferenced 3D information in a fast and efficient way, which can be used in several applications such as the 3D recording and reconstruction of complex urban areas and/or landscapes. In this study the Mobile LiDAR System is applied in the field of Cultural Heritage aiming to evaluate its performance with the purpose to document, divulgate, or to develop an architectural analysis. This study was focused on the Medieval Wall of Avila (Spain) and, specifically, the performed accuracy tests were applied in the "Alcazar" gate (National Monument from 1884). The Mobile LiDAR System is then compared to the most commonly employed active sensors (Terrestrial Laser Scanner) for large Cultural Heritage sites in regard to time, accuracy and resolution of the point cloud. The discrepancies between both technologies are established comparing directly the 3D point clouds generated, highlighting the errors affecting the architectural structures. Consequently, and based on a detailed geometrical analysis, an optimization methodology is proposed, establishing a segmented and classified cluster for the structures. Furthermore, three main clusters are settled, according to the curvature: (i) planar or low curvature; (ii) cylindrical, mild transitions and medium curvature; and (iii) the abrupt transitions of high curvature. The obtained 3D point clouds in each cluster are analyzed and optimized, considering the reference spatial sampling, according to a confidence interval and the feature curvature. The presented results suggest that Mobile LiDAR System is an optimal approach, allowing a high-speed data acquisition and providing an adequate accuracy for large Cultural Heritage sites.

show abstract

Section: Methodsmentioning

confidence: 99%

Mobile LiDAR System: New Possibilities for the Documentation and Dissemination of Large Cultural Heritage Sites

et al. 2017

View full text Add to dashboard Cite

show abstract

“…Several methods are available to detect keypoints in the literature [8,12]; the intrinsic shape signatures (ISS) method was used in this study. ISS scanned a surface and chose only points with large variations in the principal direction, which were ideal for being keypoints.…”

Section: Keypoint Extraction and Feature Descriptionmentioning

confidence: 99%

“…This study introduced a new robust surface-based automatic registration method based on 3D feature matching for point cloud registration [7][8][9][10][11][12]. The proposed method is more robust and convenient in IGNS than existing surface-based registration methods.…”

Section: Introductionmentioning

confidence: 99%

A new robust markerless method for automatic image-to-patient registration in image-guided neurosurgery system

Liu

Song

Wang

2017

Computer Assisted Surgery

View full text Add to dashboard Cite

Background: Compared with the traditional point-based registration in the image-guided neurosurgery system, surface-based registration is preferable because it does not use fiducial markers before image scanning and does not require image acquisition dedicated for navigation purposes. However, most existing surface-based registration methods must include a manual step for coarse registration, which increases the registration time and elicits some inconvenience and uncertainty. Methods: A new automatic surface-based registration method is proposed, which applies 3D surface feature description and matching algorithm to obtain point correspondences for coarse registration and uses the iterative closest point (ICP) algorithm in the last step to obtain an image-to-patient registration. Results: Both phantom and clinical data were used to execute automatic registrations and target registration error (TRE) calculated to verify the practicality and robustness of the proposed method. In phantom experiments, the registration accuracy was stable across different downsampling resolutions (18-26 mm) and different support radii (2-6 mm). In clinical experiments, the mean TREs of two patients by registering full head surfaces were 1.30 mm and 1.85 mm. Conclusion: This study introduced a new robust automatic surface-based registration method based on 3D feature matching. The method achieved sufficient registration accuracy with different real-world surface regions in phantom and clinical experiments.

show abstract

“…For example, authors in [5] tested suitability of algorithms for fall detection systems, where orientation invariance is taken into account or, as in [6], compared two keypoint detectors and descriptors using natural outdoor environment images. There are also approaches testing suitability of keypoint detection techniques for 3D applications [11][12][13].…”

Section: Related Workmentioning

confidence: 99%

“…A keypoint is said to be repeatable if the distance d between its location and the location of the nearest keypoint (in pixels) that was found after introducing a distortion to the image is less than ε [11,12,19]. Benchmark datasets often provide homography between images used in order to determine expected keypoint location.…”

Section: A Repeatabilitymentioning

confidence: 99%

Evaluation of Interest Point Detectors in Presence of Noise

Ziomek¹,

Oszust

2016

IJISA

View full text Add to dashboard Cite

Abstract-Detection of repeatable keypoints is often one of the first steps leading to obtain a solution able to recognise objects on images. Such objects are characterised by content of image patches indicated by keypoints. A given image patch is worth being described and processed in further steps, if the interest point inside of it can be found despite different image transformations or distortions. Therefore, it is important to compare keypoint detection techniques using image datasets that contain transformed or noisy images. Since most of detector evaluations rely on small datasets or are focused on a specific application of compared techniques, in this paper two large datasets which cover typical transformations, as well as challenging distortions that can occur while image processing, are used. The first dataset contains 200,000 transformed images, and it has been prepared for the purpose of this study. The second dataset, TID2013, is widely used for perceptual image quality assessment; it contains 3,000 images with 24 distortions. Finally, interest point detectors are evaluated on four datasets, and repeatability score and time of detection are used as measures of their performance.

show abstract

Performance Evaluation of 3D Keypoint Detectors

Cited by 337 publications

References 17 publications

Mobile LiDAR System: New Possibilities for the Documentation and Dissemination of Large Cultural Heritage Sites

Mobile LiDAR System: New Possibilities for the Documentation and Dissemination of Large Cultural Heritage Sites

A new robust markerless method for automatic image-to-patient registration in image-guided neurosurgery system

Evaluation of Interest Point Detectors in Presence of Noise

Contact Info

Product

Resources

About