Point-of-interest detection for range data

Viksten, Fredrik; Nordberg, Klas; Kalms, Mikael

doi:10.1109/icpr.2008.4761179

Cited by 8 publications

(5 citation statements)

References 26 publications

(28 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…, where L xx (x, σ) is the Gaussian wrapping of the second order derivative of the image I in the point X, analogously L xy (x, σ) and L yy (x, σ). Implementation of the Gaussian function for scale-space image analysis is optimal [4]. In order to track the identified characteristic points during movements and turns of the head, we estimate the Euclidean distances between those special points in the feature space, which values are present in the descriptors of special points with consideration of the known direction of movement of a part of the analyzed surface.…”

Section: Submodel Of Identification Of Characteristic Image Pointsmentioning

confidence: 99%

Multisensor Optical-Electronic Device for Calculation of Surface Parameters of the 3D Curvelinear Object

Karavaev¹,

Truphanov²

2016

Proceedings of the 7th International Conference on 3D Body Scanning Technologies, Lugano, Switzerland, 30 Nov.-1 Dec. 2016

View full text Add to dashboard Cite

The research covers the operation principles and structure functional organization of specialized optical-electronic device for calculation of a volumetric object. The purpose of the device is to build the 3D plane, which with the given degree of precision approximates the surface of the top and nape sections of a human head intended for implication in the production of customized headwear and elements of protectiveclothing or for other similar tasks related to the analysis of the 3D objectshape. The proposed device provides the ability to build the 3D model of an object under study by acquiring and analyzing the series of images of its surface continuously acquired from three or four optical-electronic sensors throughout the measurement process. The number of the optical-electronic sensors depends on the configuration of the optical-electronic device. Optical-electronic sensors organized in a given order relative to each other constitute a unified system, which in its turn is to be placed in the prescribed range of positions in relation to the head. In order to simplify the procedure and enhance the accuracy of calculation of special coordinates for a set of points of the object surface and reduce the computational complexity of the developed algorithms we use the tight-fitting contact hat consisting of a set of special contrast markers., which is to be put on the head. The following beneficial features of the proposed method and the optical-electronic device have the following beneficial features: (i) no strict requirements to interrelated positioning of the analyzed object and the system of the optical-electronic sensors; (ii) possibility to use commercially available web cameras as optical-electronic sensors; and (iii) no structurally complex and moving mechanical elements, which allow its mass use with no reduction in the accuracy of calculated point coordinates on the surface of an object required for practical tasks. The experiments we conducted showed that depending on the interrelated position of the device and the analyzed object our optical-electronic device gives the calculation error of the 3D coordinates of about ±2 to ±7 mm, which is sufficient for practical tasks.

show abstract

Section: Submodel Of Identification Of Characteristic Image Pointsmentioning

confidence: 99%

Multisensor Optical-Electronic Device for Calculation of Surface Parameters of the 3D Curvelinear Object

Karavaev¹,

Truphanov²

2016

Proceedings of the 7th International Conference on 3D Body Scanning Technologies, Lugano, Switzerland, 30 Nov.-1 Dec. 2016

View full text Add to dashboard Cite

show abstract

“…Hartkens et al 14 made a comparison of nine di®erential detectors and found that point of interest detectors based on the Harris matrix performs better than the ones based on measures of curvature. Viksten et al 32 used 3D Harris over a 2D range image.…”

Section: Literature Review and Related Workmentioning

confidence: 99%

Interest Point Detection in 3D Point Cloud Data Using 3D Sobel-Harris Operator

Hafiz

Youssef

Sheta

et al. 2015

Int. J. Patt. Recogn. Artif. Intell.

View full text Add to dashboard Cite

Manual selection of features from massive unstructured point cloud data is a very time-consuming task that requires a considerable amount of human intervention. This work is motivated by the need of fast and simple algorithm to obtain robust, stable and well-localized interest points that are used for subsequent processing in computer vision real-time applications. This paper presents an algorithm for detection of interest points in three-dimensional (3D) point cloud data by using a combined 3D Sobel-Harris operator. The proposed algorithm is compared with six state-of-the-art approaches used to identify the true feature points. Extensive experiments were carried out using synthetic benchmark and real datasets. The datasets were selected with di®erent sizes, features and scales. The results were evaluated against human generated ground truth and prede¯ned feature points. Three measures were used to evaluate the algorithm accuracy, namely localization accuracy L e , False Positive Error (FPE) and False Negative Errors (FNE). Also, the complexity analysis of the proposed algorithm is presented. The results show that the proposed algorithm can identify the interest points with accepted accuracy. It works directly on point cloud datasets and shows superiority when compared with other methods work on 3D mesh data.

show abstract

“…Because the 3D models are influenced by the type of the 3D model acquisition devices, the position and pose of the 3D models, the 3D keypoint detection technique is still not mature in theory and application. Therefore, it is essential to study the keypoint detection algorithm of the 3D model [6,9,18,19].…”

Section: Introductionmentioning

confidence: 99%

Robust 3D keypoint detection method based on double Gaussian weighted dissimilarity measure

Zeng

Wang

Dong

2016

Multimed Tools Appl

View full text Add to dashboard Cite

This paper proposes a novel multiscale 3D keypoint detection method via the double Gaussian weighted dissimilarity measure. At each scale, the shape index value and the double Gaussian weighted dissimilarity measure value of each 3D point are firstly computed. Then the candidate keypoints with local maximum dissimilarity measure values are determined. Finally the final 3D keypoints are detected under our proposed multiscale detection scheme. As the dissimilarity measure used in this paper has better robust descriptive ability and is rotation and translation transformation invariant, the proposed detection method is robust to noise, rotation and translation transformation. Extensive experimental results have shown that using our proposed multiscale detection method, we can detect the keypoints with higher repeatability under different noise levels.

show abstract

Point-of-interest detection for range data

Cited by 8 publications

References 26 publications

Multisensor Optical-Electronic Device for Calculation of Surface Parameters of the 3D Curvelinear Object

Multisensor Optical-Electronic Device for Calculation of Surface Parameters of the 3D Curvelinear Object

Interest Point Detection in 3D Point Cloud Data Using 3D Sobel-Harris Operator

Robust 3D keypoint detection method based on double Gaussian weighted dissimilarity measure

Contact Info

Product

Resources

About