Fusing points and lines for high performance tracking

Rosten, Edward; Drummond, Tom

doi:10.1109/iccv.2005.104

Cited by 923 publications

(565 citation statements)

References 15 publications

Supporting

Mentioning

525

Contrasting

Unclassified

Order By: Relevance

“…Firstly, the possibilities is overviewed that OpenCV can give about feature tracking. The currently supported feature detectors in OpenCV are as follows: AGAST (Mair et al, 2010), AKAZE (Pablo Alcantarilla (Georgia Institute of Technolog), 2013), BRISK (Leutenegger et al, 2011), FAST (Rosten and Drummond, 2005), GFTT (Tomasi, C. and Shi, J., 1994) (Good Features To Track -also known as Shi-Tomasi corners), KAZE (Alcantarilla et al, 2012), MSER (Matas et al, 2002), ORB (Rublee et al, 2011). However, if one compiles the contrib(nonfree) repository with the OpenCV, the following detectors can also be used: SIFT (Lowe, 1999b), STAR (Agrawal and Konolige, 2008), and SURF (Bay et al, 2008).…”

Section: Comparison Of Feature Matchers Implemented In Opencv3mentioning

confidence: 99%

Ground-Truth Tracking Data Generation Using Rotating Real-World Objects

Pusztai

Hajder

2017

Communications in Computer and Information Science

View full text Add to dashboard Cite

Abstract. Quantitative comparison of feature matchers/trackers is essential in 3D computer vision as the accuracy of spatial algorithms mainly depends on the quality of feature matching. This paper shows how a structured-light applying turntable-based evaluation system can be developed. The key problem here is the highly accurate calibration of scanner components. The ground truth (GT) tracking data generation is carried out for seven testing objects. It is shown how the OpenCV3 feature matchers can be compared on our GT data, and the obtained quantitative results are also discussed in detail.

show abstract

Section: Comparison Of Feature Matchers Implemented In Opencv3mentioning

confidence: 99%

Ground-Truth Tracking Data Generation Using Rotating Real-World Objects

Pusztai

Hajder

2017

Communications in Computer and Information Science

View full text Add to dashboard Cite

show abstract

“…Focus points are considered to be features of the color image; i.e., points that are most likely of visual interest such as edges or corners [Hub88]. Therefore, we perform corner detection on the previously obtained color data by using the FAST feature detector [RD05]. We then select a single feature (per time step) that we consider to be the most likely focus point and set the focal plane according to the depth value of the chosen feature.…”

Section: Scene Rendering and Focus Computationmentioning

confidence: 99%

Real-time Simulation of Human Vision using Temporal Compositing with CUDA on the GPU

Nießner¹,

Kuhnert²,

Selgrad³

et al. 2013

PARS-Mitteilungen

View full text Add to dashboard Cite

We present a novel approach that simulates human vision including visual defects such as glaucoma by temporal composition of human vision in real-time on the GPU. Therefore, we determine eye focus points every time step and adapt the lens accommodation of our virtual eye model accordingly. The focal distance is then used to determine bluriness of observed scene regions; i.e., we compute defocus for all visible pixels. In order to simulate the visual memory we introduce a sharpness field where we integrate defocus values temporally. That allows for memorizing sharply perceived scene points. For visualization, we ray trace the virtual scene environment while incorporating depth of field based on the sharpness field data. Thus, our algorithm facilitates the simulation of human vision mimicing the visual memory. We consider this to be particularly useful for illustration purposes for patients with visual defects such as glaucoma. In order to run our algorithm in real-time we employ massively parallel graphics hardware.

show abstract

“…Even though this deterministic approach might seem simplistic, we prefer it because, in practice, it is very hard to evaluate a new radius from currently valid matches. For example, a similar problem is solved in Rosten and Drummond (2005) using an expectation minimization (EM) approach and the authors have to "give a kick downwards" to their blurring factor when EM converges too early. Here we show that value of η has relatively little impact on the optimizer's behavior.…”

Section: Parameters and Initial Conditionsmentioning

confidence: 99%

“…Rigid object detection and tracking have been extensively studied and effective, robust, and real-time solutions proposed Lowe, 2004;Rosten and Drummond, 2005). The two are of course complementary since trackers require initialization and, no matter how good they may be, will sometimes lose track, for example, because of severe occlusions.…”

Section: Introductionmentioning

confidence: 99%

Fast Non-Rigid Surface Detection, Registration and Realistic Augmentation

2007

View full text Add to dashboard Cite

Abstract. We present a real-time method for detecting deformable surfaces, with no need whatsoever for a priori pose knowledge.Our method starts from a set of wide baseline point matches between an undeformed image of the object and the image in which it is to be detected. The matches are used not only to detect but also to compute a precise mapping from one to the other. The algorithm is robust to large deformations, lighting changes, motion blur, and occlusions. It runs at 10 frames per second on a 2.8 GHz PC. We demonstrate its applicability by using it to realistically modify the texture of a deforming surface and to handle complex illumination effects.Combining deformable meshes with a well designed robust estimator is key to dealing with the large number of parameters involved in modeling deformable surfaces and rejecting erroneous matches for error rates of more than 90%, which is considerably more than what is required in practice.

show abstract

Fusing points and lines for high performance tracking

Cited by 923 publications

References 15 publications

Ground-Truth Tracking Data Generation Using Rotating Real-World Objects

Ground-Truth Tracking Data Generation Using Rotating Real-World Objects

Real-time Simulation of Human Vision using Temporal Compositing with CUDA on the GPU

Fast Non-Rigid Surface Detection, Registration and Realistic Augmentation

Contact Info

Product

Resources

About