Model-Based Pose Estimation for Rigid Objects

Lourakis, Manolis; Zabulis, Xenophon

doi:10.1007/978-3-642-39402-7_9

Cited by 44 publications

(39 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Specifically, the detection of particular markers in an image combined with the knowledge of their physical location on the turntable provided a set of 2D-3D correspondences. These were used to estimate the camera pose in the turntable coordinate system by robustly solving the PnP problem and then refining the estimated 6D pose by non-linearly minimizing the cumulative re-projection error with the posest library from [31]. The root mean square re-projection error, which was calculated at marker corners in all test images, is 1.27 px for Carmine, 1.37 px for Kinect, and 1.50 px for Canon.…”

Section: Calibration Of Sensorsmentioning

confidence: 99%

T-LESS: An RGB-D Dataset for 6D Pose Estimation of Texture-Less Objects

Hodaň

Haluza

Obdržálek

et al. 2017

2017 IEEE Winter Conference on Applications of Computer Vision (WACV)

Self Cite

377

329

View full text Add to dashboard Cite

We introduce T-LESS, a new public dataset for estimating the 6D pose, i.e. translation and rotation, of texture-less rigid objects. The dataset features thirty industry-relevant objects with no significant texture and no discriminative color or reflectance properties. The objects exhibit symmetries and mutual similarities in shape and/or size. Compared to other datasets, a unique property is that some of the objects are parts of others. The dataset includes training and test images that were captured with three synchronized sensors, specifically a structured-light and a time-of-flight RGB-D sensor and a high-resolution RGB camera. There are approximately 39K training and 10K test images from each sensor. Additionally, two types of 3D models are provided for each object, i.e. a manually created CAD model and a semi-automatically reconstructed one. Training images depict individual objects against a black background. Test images originate from twenty test scenes having varying complexity, which increases from simple scenes with several isolated objects to very challenging ones with multiple instances of several objects and with a high amount of clutter and occlusion. The images were captured from a systematically sampled view sphere around the object/scene, and are annotated with accurate ground truth 6D poses of all modeled objects. Initial evaluation results indicate that the state of the art in 6D object pose estimation has ample room for improvement, especially in difficult cases with significant occlusion. The T-LESS dataset is available online at cmp.felk.cvut.cz/t-less. The visual appearance of a texture-less object is dominated by its global shape, color, reflectance properties, and the configuration of light sources. The lack of texture implies that the object cannot be reliably recognized with traditional techniques relying on photometric local patch detectors and descriptors [9,31]. Instead, recent approaches that can deal with texture-less objects have focused on local 3D feature description [33,51,19], and semi-global or arXiv:1701.05498v1 [cs.CV]

show abstract

Section: Calibration Of Sensorsmentioning

confidence: 99%

T-LESS: An RGB-D Dataset for 6D Pose Estimation of Texture-Less Objects

Hodaň

Haluza

Obdržálek

et al. 2017

2017 IEEE Winter Conference on Applications of Computer Vision (WACV)

Self Cite

377

329

View full text Add to dashboard Cite

show abstract

“…This problem, also known as the PnP or camera resection problem, 8 has received much attention due to is applicability in various domains. Exterior orientation is typically dealt with by embedding minimalsize PnP solvers to robust regression frameworks such as RANSAC (see [38] and references therein). However, as minimal solutions ignore much of the redundancy present in the data, they suffer from inaccuracies.…”

Section: Exterior Orientationmentioning

confidence: 99%

“…However, as minimal solutions ignore much of the redundancy present in the data, they suffer from inaccuracies. To remedy this, an additional step comprised of nonlinear optimization with the Levenberg-Marquardt algorithm is employed to minimize the reprojection error pertaining to all inliers [38].…”

Section: Exterior Orientationmentioning

confidence: 99%

“…8.3, we extracted the 3D points projecting to their first frames. Then, those 3D points along with their projections in the first frame of each dataset were used for estimating the corresponding camera poses using the posest library implementing [38]. We also modified the nonlinear refinement step of posest to employ a rotation parameterization based on MRPs and compared it with its native axis-angle parameterization.…”

Section: Exterior Orientationmentioning

confidence: 99%

See 1 more Smart Citation

Modified Rodrigues Parameters: An Efficient Representation of Orientation in 3D Vision and Graphics

2017

Self Cite

View full text Add to dashboard Cite

Modified Rodrigues parameters (MRPs) are triplets in R 3 bijectively and rationally mapped to quaternions through stereographic projection. We present here a compelling case for MRPs as a minimal degree-of-freedom parameterization of orientation through novel solutions to prominent problems in the fields of 3D vision and computer graphics. In our primary contribution, we show that the derivatives of a unit quaternion in terms of its MRPs are simple polynomial expressions of its scalar and vector part. Furthermore, we show that updates to unit quaternions from perturbations in parameter space can be computed without explicitly invoking the parameters in the computations. Based on the former, we introduce a novel approach for designing orientation splines by configuring their backprojections in 3D space. Finally, in the general topic of nonlinear optimization for geometric vision, we run performance analyses and provide comparisons on the convergence behavior of MRP parameterizations on the tasks of absolute orientation, exterior orientation and large-scale bundle adjustment of public datasets.

show abstract

“…The minimization is made more immune to noise caused by mislocalized image points by substituting the squared reprojection error with a robust cost function (i.e., M-estimator). Our pose estimation approach is detailed in [10].…”

Section: Scale Estimation From Known Object Motionmentioning

confidence: 99%

Accurate Scale Factor Estimation in 3D Reconstruction

Lourakis

Zabulis

2013

Computer Analysis of Images and Patterns

Self Cite

View full text Add to dashboard Cite

Abstract.A well-known ambiguity in monocular structure from motion estimation is that 3D reconstruction is possible up to a similarity transformation, i.e. an isometry composed with isotropic scaling. To raise this ambiguity, it is commonly suggested to manually measure an absolute distance in the environment and then use it to scale a reconstruction accordingly. In practice, however, it is often the case that such a measurement cannot be performed with sufficient accuracy, compromising certain uses of a 3D reconstruction that require the acquisition of true Euclidean measurements. This paper studies three alternative techniques for obtaining estimates of the scale pertaining to a reconstruction and compares them experimentally with the aid of real and synthetic data.

show abstract

Model-Based Pose Estimation for Rigid Objects

Cited by 44 publications

References 20 publications

T-LESS: An RGB-D Dataset for 6D Pose Estimation of Texture-Less Objects

T-LESS: An RGB-D Dataset for 6D Pose Estimation of Texture-Less Objects

Modified Rodrigues Parameters: An Efficient Representation of Orientation in 3D Vision and Graphics

Accurate Scale Factor Estimation in 3D Reconstruction

Contact Info

Product

Resources

About