Pose Estimation of Known Objects by Efficient Silhouette Matching

Reinbacher, Christian; Rüther, Matthias; Bischof, Horst

doi:10.1109/icpr.2010.270

Cited by 20 publications

(24 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Specifically, different simulations are characterized by random extractions for the range and pointing uncertainties, as well as for the detection process and the presence of outliers based on the computed values for the probability of detection and on the percentage of outliers. Regarding the pose initialization, a large angular step (Δ¼ 201) has been adopted in the TM algorithm to sample the attitude parameter space, thus limiting the number of templates to about 3600, much less than typical values which range between 5000 and 10,000 [17]. This allows an initial pose estimation accuracy of order Δ in the relative attitude, and of about 1 m in the relative position.…”

Section: Resultsmentioning

confidence: 99%

“…A fundamental limitation of this technique is the high computational cost, since pose search must be carried out over the entire pose space that has to be densely sampled in order to get sufficient accuracy [16]. Hence, further studies are needed to overcome this limitation starting from the possibility of organizing the pose database with hierarchical structures [17], generate the database on line (i.e. after target first acquisition) and operating with sparse point clouds.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Uncooperative pose estimation with a LIDAR-based system

et al. 2015

View full text Add to dashboard Cite

Section: Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Uncooperative pose estimation with a LIDAR-based system

et al. 2015

View full text Add to dashboard Cite

“…2D methods can be found in (Haniff et al, 2011), (Xu et al, 2008) or (Fan et al, 2014) and are, in principle, suitable to achieve sub-pixel accuracies. 3D approaches determining the pose of an object like reported in (Ulrich et al, 2009) or (Reinbacher et al, 2010) have advantages over 2D methods but may lead to insufficient results according to real time capability or accuracy for specific tasks. Cars need to pass several alignment and initial operation processes before their delivery.…”

Section: Related Workmentioning

confidence: 99%

Determination of Steering Wheel Angles During Car Alignment by Image Analysis Methods

Mueller

Voegtle

2016

Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci.

View full text Add to dashboard Cite

ABSTRACT:Optical systems for automatic visual inspections are of increasing importance in the field of automation in the industrial domain. A new application is the determination of steering wheel angles during wheel track setting of the final inspection of car manufacturing. The camera has to be positioned outside the car to avoid interruptions of the processes and therefore, oblique images of the steering wheel must be acquired. Three different approaches of computer vision are considered in this paper, i.e. a 2D shape-based matching (by means of a plane to plane rectification of the oblique images and detection of a shape model with a particular rotation), a 3D shape-based matching approach (by means of a series of different perspectives of the spatial shape of the steering wheel derived from a CAD design model) and a point-to-point matching (by means of the extraction of significant elements (e.g. multifunctional buttons) of a steering wheel and a pairwise connection of these points to straight lines). The HALCON system (HALCON, 2016) was used for all software developments and necessary adaptions. As reference a mechanical balance with an accuracy of 0.1° was used. The quality assessment was based on two different approaches, a laboratory test and a test during production process. In the laboratory a standard deviation of ±0.035° (2D shape-based matching), ±0.12° (3D approach) and ±0.029° (point-to-point matching) could be obtained. The field test of 291 measurements (27 cars with varying poses and angles of the steering wheel) results in a detection rate of 100% and ±0.48° (2D matching) and ±0.24° (point-to-point matching). Both methods also fulfil the request of real time processing (three measurements per second).

show abstract

“…Liebelt and Schertler [21] proposed a new similarity measure that combines the contour matching and the appearance-based mutual information measure in order to increase alignment precision. In [12], the normalized cross correlation is used to evaluate the similarity between the two contours and the contour matching is accelerated by a hierarchical clustering scheme. Nevertheless, this kind of registration methods is strongly dependent on foreground segmentation technique and is sensitive to occlusion.…”

Section: Related Workmentioning

confidence: 99%

“…In recent years, 3D-2D vehicle registration has attracted more and more attentions, which provides a new way for vehicle recognition [8][9][10][11], localization [12][13][14] and tracking [15][16][17][18]. With the rapid development of 3D modeling technology, 3D vehicle model can be easily obtained.…”

Section: Introductionmentioning

confidence: 99%