Klaus Denker's Publications

2015

On-line CAD Reconstruction with Accumulated Means of Local Geometric Properties

On-line CAD Reconstruction with Accumulated Means of Local Geometric Properties.
K. Denker, B. Hamann, G. Umlauf
In: J.-D. Boissonnat et al. (eds), Curves and Surfaces, Springer, pp. 181–201, 2015.

Reconstruction of hand-held laser scanner data is used in industry primarily for reverse engineering. Traditionally, scanning and reconstruction are separate steps. The operator of the laser scanner has no feedback from the reconstruction results. On-line reconstruction of the CAD geometry allows for such an immediate feedback.

We propose a method for on-line segmentation and reconstruction of CAD geometry from a stream of point data based on means that are updated on-line. These means are combined to define complex local ge- ometric properties, e.g., to radii and center points of spherical regions. Using means of local scores, planar, cylindrical, and spherical segments are detected and extended robustly with region growing. For the on-line computation of the means we use so-called accumulated means. They allow for on-line insertion and removal of values and merging of means.

Our results show that this approach can be performed on-line and is robust to noise. We demonstrate that our method reconstructs spherical, cylindrical, and planar segments on real scan data containing typical errors caused by hand-held laser scanners.

Download PDF
The final publication is available at link.springer.com.
DOI: 10.1007/978-3-319-22804-4_14
Bibtex

Support Vector Machines for Classification of Geometric Primitives in Point Clouds.

Support Vector Machines for Classification of Geometric Primitives in Point Clouds.
M. Caputo, K. Denker, M. O. Franz, P. Laube, G. Umlauf
In: J.-D. Boissonnat et al. (eds), Curves and Surfaces, Springer, pp. 80–95, 2015.

Classification of point clouds by different types of geometric primitives is an essential part in the reconstruction process of CAD geometry. We use support vector machines (SVM) to label patches in point clouds with the class labels tori, ellipsoids, spheres, cones, cylinders or planes. For the classification features based on different geometric properties like point normals, angles, and principal curvatures are used. These geometric features are estimated in the local neighborhood of a point of the point cloud. Computing these geometric features for a random subset of the point cloud yields a feature distribution. Different features are combined for achieving best classification results. To minimize the time consuming training phase of SVMs, the geometric features are first evaluated using linear discriminant analysis (LDA).

LDA and SVM are machine learning approaches that require an initial training phase to allow for a subsequent automatic classification of a new data set. For the training phase point clouds are generated using a simulation of a laser scanning device. Additional noise based on an laser scanner error model is added to the point clouds. The resulting LDA and SVM classifiers are then used to classify geometric primitives in simulated and real laser scanned point clouds.

Compared to other approaches, where all known features are used for classification, we explicitly compare novel against known geometric features to prove their effectiveness.

Download PDF
The final publication is available at link.springer.com.
DOI: 10.1007/978-3-319-22804-4_7
Bibtex

2014

Acquisition and On-line Reconstruction of 3D Point Data from Hand-held Laser Scanners and Multi-camera Stereo-matching

Acquisition and On-line Reconstruction of 3D Point Data from Hand-held Laser Scanners and Multi-camera Stereo-matching.
Klaus Denker
PhD Thesis: University of Kaiserslautern, 2014.
Advisors: Hans Hagen, Bernd Hamann, Georg Umlauf

Three dimensional (3d) point data is used in industry for measurement and reverse engineering. Precise point data is usually acquired with triangulating laser scanners or high precision structured light scanners. Lower precision point data is acquired by real-time structured light devices or by stereo matching with multiple cameras. The basic principle of all these methods is the so-called triangulation of 3d coordinates from two dimensional (2d) camera images.

This dissertation contributes a method for multi-camera stereo matching that uses a system of four synchronized cameras. A GPU based stereo matching method is presented to achieve a high quality reconstruction at interactive frame rates. Good depth resolution is achieved by allowing large disparities between the images. A multi level approach on the GPU allows a fast processing of these large disparities. In reverse engineering, hand-held laser scanners are used for the scanning of complex shaped objects. The operator of the scanner can scan complex regions slower, multiple times, or from multiple angles to achieve a higher point density. Traditionally, computer aided design (CAD) geometry is reconstructed in a separate step after the scanning. Errors or missing parts in the scan prevent a successful reconstruction. The contribution of this dissertation is an on-line algorithm that allows the reconstruction during the scanning of an object. Scanned points are added to the reconstruction and improve it on-line. The operator can detect the areas in the scan where the reconstruction needs additional data.

First, the point data is thinned out using an octree based data structure. Local normals and principal curvatures are estimated for the reduced set of points. These local geometric values are used for segmentation using a region growing approach. Implicit quadrics are fitted to these segments. The canonical form of the quadrics provides the parameters of basic geometric primitives.

An improved approach uses so called accumulated means of local geometric properties to perform segmentation and primitive reconstruction in a single step. Local geometric values can be added and removed on-line to these means to get a stable estimate over a complete segment. By estimating the shape of the segment it is decided which local areas are added to a segment. An accumulated score estimates the probability for a segment to belong to a certain type of geometric primitive. A boundary around the segment is reconstructed using a growing algorithm that ensures that the boundary is closed and avoids self intersections.

Download at kluedo.ub.uni-kl.de.
Bibtex

Learning geometric primitives in point clouds

Learning geometric primitives in point clouds.
M. Caputo, K. Denker, M. O. Franz, P. Laube, G. Umlauf
In: Symposium on Geometry Processing 2014, EuroGraphics, extended Abstract.

Primitive recognition in 3D point clouds is an important aspect in reverse engineering. We propose a method for primitive recognition based on machine learning approaches. The machine learning approaches used for the classification are linear discriminant analysis (LDA) and multi-class support vector machines (SVM). For the classification process local geometric properties (features) of the point cloud are computed based on point relations, normals, and principal curvatures. For the training phase point clouds are generated using a simulation of a laser scanning device based on ray tracing with an error model. The classification rates of novel, curvature based geometric features are compared to known geometric features to prove the effectiveness of the approach.

Download PDF
Bibtex

2013

On-line Reconstruction of CAD Geometry

On-line Reconstruction of CAD Geometry.
K. Denker, D. Hagel, J. Raible, G. Umlauf, B. Hamann
In: Proceedings of the International Conference on 3D Vision, IEEE, pp. 151–158, 2013.

In reverse engineering and computer-aided design (CAD) applications point cloud data is usually manually scanned, reconstructed, and post-processed in separated steps. When point cloud data resulting from a scanning process do not satisfy certain necessary reconstruction requirements, one must perform scanning again to enable proper reconstruction. On-line reconstruction of 3d geometry allows one to generate and update a CAD reconstruction on-line during the scanning process with an hand-held laser scanner. Thus, regions where the scanned data is insufficient for the reconstruction are detected on the fly to allow an immediate correction and improvement of the scanned data. This enables the operator to focus on critical regions in the scanned data to improve the reconstruction quality.

We present an on-line segmentation and on-line reconstruc- tion of basic geometric primitives. The presented methods allow for a real-time processing of a point stream. They utilize data structures that can be updated at any time when additional data from the stream has to be processed. This data is used to complete and improve the segmentation and reconstruction during the scanning process.

Download PDF
The final publication is available at ieeexplore.ieee.org.
DOI: 10.1109/3DV.2013.28
Bibtex

2012

3d hand gesture recognition

3D Hand Gesture Recognition Based on Sensor Fusion of Commodity Hardware.
M. Caputo, K. Denker, B. Dums, G. Umlauf
In: H. Reiterer, O. Deussen (eds.), Mensch & Computer 2012, Oldenbourg, pp. 293–302, 2012.

With the advent of various video game consoles and tablet devices gesture recognition got quite popular to control computer systems. E.g. touch screens allow for an intuitive control of small 2d user interfaces with finger gestures. For interactive manipulation of 3d objects in a large 3d projection environment a similar intuitive 3d interaction method is required. In this paper, we present a dynamic 3d hand and arm gesture recognition system using commodity hardware. The input data is captured with low cost depth sensors (e.g. Microsoft Kinect) and HD color sensors (e.g. Logitech C910). Our method combines dynamic hand and arm gesture recognition based on the depth sensor with static hand gesture recognition based on the HD color sensor.

Download PDF
The final publication is available at www.oldenbourg-link.com.
DOI: 10.1524/9783486718782.293
Bibtex

hand-held laser scanner

A Hand-held Laser Scanner based on Multi-camera Stereo-matching.
C. Bender, K. Denker, M. Friedrich, K. Hirt, G. Umlauf
In: C. Garth et al. (eds.), VLUDS 2011, OpenAccess Series in Informatics (OASIcs), 27: 123–133, 2012.

Most laser scanners in engineering are extended versions of tactile measuring machines. These high precision devices are typically very expensive and hardware modifications are not possible without impairing the precision of the device.

For these reasons we built our own laser-scanner system. It is based on a multi-camera reconstruction system developed for fast 3D face reconstructions. Based on this camera system, we developed a laser-scanner using GPU accelerated stereo-matching techniques and a hand-held line-laser probe. The resulting reconstruction is solely based on the known camera positions and parameters. Thus, it is not necessary to track the position and movement of the line-laser probe. This yields an inexpensive laser-scanner system where every hardware component can be modified individually for experiments and future extensions of the system.

Download PDF
Proceedings
DOI: 10.4230/OASIcs.VLUDS.2011.123
Bibtex

2011

hybrid face recognition

Hybrid Face Recognition Based on Real-Time Multi-Camera Stereo-Matching.
J. Hensler, K. Denker, M. O. Franz, G. Umlauf
In: G. Bebis et al. (eds.), ISVC 2011, Lecture Notes in Computer Science (LNCS) 6939, pp. 158–167, 2011.

Multi-camera systems and GPU-based stereo-matching methods allow for a real-time 3d reconstruction of faces. We use the data generated by such a 3d reconstruction for a hybrid face recognition system based on color, accuracy, and depth information. This system is structured in two subsequent phases: geometry-based data preparation and face recognition using wavelets and the AdaBoost algorithm. It requires only one reference image per person. On a data base of 500 recordings, our system achieved detection rates ranging from 95% to 97% with a false detection rate of 2% to 3%. The computation of the whole process takes around 1.1 seconds.

Download PDF
The final publication is available at www.springerlink.com.
DOI: 10.1007/978-3-642-24031-7_16
Bibtex

multi-camera matching

Accurate real-time multi-camera stereo-matching on the GPU for 3d reconstruction.
K. Denker, G. Umlauf
In: Journal of WSCG, 19(1): 9–16, 2011.

Using multi-camera matching techniques for 3d reconstruction there is usually the trade-off between the quality of the computed depth map and the speed of the computations. Whereas high quality matching methods take several seconds to several minutes to compute a depth map for one set of images, real-time methods achieve only low quality results. In this paper we present a multi-camera matching method that runs in real-time and yields high resolution depth maps.

Our method is based on a novel multi-level combination of normalized cross correlation, deformed matching windows based on the multi-level depth map information, and sub-pixel precise disparity maps. The whole process is implemented completely on the GPU. With this approach we can process four 0.7 megapixel images in 129 milliseconds to a full resolution 3d depth map. Our technique is tailored for the recognition of non-technical shapes, because our target application is face recognition.

Download PDF
Proceedings
Bibtex

survey on benchmarks

Survey on benchmarks for a GPU based multi camera stereo matching algorithm.
K. Denker, G. Umlauf
In: A. Middel et al. (eds.), VLUDS 2010, OpenAccess Series in Informatics (OASIcs), 19: 20–26, 2011.

Stereo matching algorithms and multi camera reconstruction algorithms are usually compared using benchmarks. These benchmarks compare the quality of the resulting depth map or reconstructed surface mesh. We describe the differences between several known stereo and multi-view stereo benchmarks and their various datasets. Also the modifications that are necessary to use our own GPU based multi camera stereo matching algorithm with the data from these benchmarks are discussed.

Download PDF
Proceedings
DOI: 10.4230/OASIcs.VLUDS.2010.20
Bibtex

realtimetriangulation

Real-time triangulation of point streams.
K. Denker, B. Lehner, G. Umlauf
In: Engineering with Computers, 27(1): 67–80, 2011

Hand-held laser scanners are commonly used in industry for reverse engineering and quality measurements. In this process, it is difficult for the human operator to scan the target object completely and uniformly. Therefore, an interactive triangulation of the scanned points can assist the operator in this task.

In this paper we describe the technical and implementational details of our real-time triangulation approach for point streams, presented at the 17th International Meshing Roundtable. Our method computes a triangulation of the point stream generated by the laser scanner online, i.e., the data points are added to the triangulation as they are received from the scanner. Multiple scanned areas and areas with a higher point density result in a finer mesh and a higher accuracy. On the other hand, the vertex density adapts to the estimated surface curvature. To guide the operator the resulting triangulation is rendered with a visualization of its uncertainty and the display of an optimal scanning direction.

Download PDF
The final publication is available at www.springerlink.com.
DOI: 10.1007/s00366-010-0181-y
Bibtex

2008

onlinetriangulation

Online triangulation of laser-scan data.
K. Denker, B. Lehner, G. Umlauf
In: R. Garimella (ed.), Proceedings of the 17th International Meshing Roundtable, Springer, pp. 415–432, 2008.

Hand-held laser scanners are used massively in industry for reverse engineering and quality measurements. In this process, it is difficult for the human operator to cover the scanned object completely and uniformly. Therefore, an interactive triangulation of the scanned surface points can assist the human operator in this task.

Our method computes a triangulation of the point stream generated by the laser scanner online, i.e., the data points are added to the triangulation as they are received from the scanner. Multiple scanned areas and areas with a higher point density result in a finer mesh and a higher accuracy. On the other hand, the vertex density adapts to the estimated surface curvature. To assist the human operator the resulting triangulation is rendered with a visualization of its faithfulness. Additionally, our triangulation method allows for a level-of-detail representation to reduce the mesh complexity for fast rendering on low-cost graphics hardware.

Download PDF
The final publication is available at www.springerlink.com.
DOI: 10.1007/978-3-540-87921-3_25
Bibtex
Video

back