Time filter

Source Type

Waardenburg, Netherlands

Wijnhoven R.G.J.,ViNotion | Wijnhoven R.G.J.,TU Eindhoven | De With P.H.N.,TU Eindhoven | De With P.H.N.,CycloMedia Technology
Proceedings of the IEEE International Conference on Computer Vision

We present a novel algorithm for unsupervised subcategorization of an object class, in the context of object detection. Dividing the detection problem into smaller subproblems simplifies the object vs. background classification. © 2011 IEEE. Source

Han J.,TU Eindhoven | De With P.H.N.,TU Eindhoven | De With P.H.N.,CycloMedia Technology
Multimedia Tools and Applications

This paper aims at generating an automated way to evaluate the team-behavior of trainees in a delivery simulation course using video-processing techniques with emphasis on multiple people tracking. The paper is composed of two interacting, but clearly separated stages: moving people segmentation and multiple people tracking. At people segmentation stage, the combination of the Gaussian Mixture Model (GMM) and the Dynamic Markov Random Fields (DMRF) technique helps to extract the foreground pixels. For a better extraction of the human silhouettes, the energy function of DMRF is extended with texture information. At multiple people tracking stage, we concentrate on solving human-occlusion problem caused by interacting persons based on silhouette data and a non-linear regression model. Our model effectively transfers the person location problem during the occlusion into the finding of the local maximum points on a smooth curve, so that visual persons in the partial or complete occlusion can still be precisely captured. We have compared our algorithm with two other popular tracking algorithms: mean-shift and particle-filter. Experimental results reveal that the correctness of our method is much higher than the mean-shift algorithm and slightly lower than a particle-filter, however, with the major benefit of being a factor of 10-15 faster in computing. © 2009 The Author(s). Source

Alsadik B.,CycloMedia Technology
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences - ISPRS Archives

The ongoing technical improvements in photogrammetry, Geomatics, computer vision (CV), and robotics offer new possibilities for many applications requiring efficient acquisition of three-dimensional data. Image orientation is one of these important techniques in many applications like mapping, precise measurements, 3D modeling and navigation. Image orientation comprises three main techniques of resection, intersection (triangulation) and relative orientation, which are conventionally solved by collinearity equations or by using projection and fundamental matrices. However, different problems still exist in the state - of -the -art of image orientation because of the nonlinearity and the sensitivity to proper initialization and spatial distribution of the points. In this research, a modified method is presented to solve the triangulation problem using inclined angles derived from the measured image coordinates and based on spherical trigonometry rules and vector geometry. The developed procedure shows promising results compared to collinearity approach and to converge to the global minimum even when starting from far approximations. This is based on the strong geometric constraint offered by the inclined angles that are enclosed between the object points and the camera stations. Numerical evaluations with perspective and panoramic images are presented and compared with the conventional solution of collinearity equations. The results show the efficiency of the developed model and the convergence of the solution to global minimum even with improper starting values. Source

Method for producing a digital photo comprising pixels, wherein at least some of the pixels comprise position information, the method comprising of: taking a photo with a known geometry of the optical system with which the photo is taken; recording the position from which the photo has been taken; recording the direction in which the photo has been taken; providing a three-dimensional model comprising points which are comprised by at least an outer surface of an object in the field of view of the position where the photo has been taken, wherein the points comprise position information; relating a pixel on the one hand and a corresponding point in the three-dimensional model on the other; and recording, in a manner associated with the pixel, the position information of the corresponding pixel.

The present invention relates to an apparatus and method for detecting and recognizing an object in an image, the method comprising a plurality of stages, wherein at least one of the stages comprises an integrated approach of feature detection and object recognition of at least a part of the object. In a further embodiment at least one of the stages comprises identifying an image part that contains a feature point, and matching the image part to a set of hierarchies of templates, wherein a hierarchy comprises templates for an object to recognized, a template describes at least a part of an object to be recognized, and a child template describes a sub-part of the part of the object described by its parent template.

Discover hidden collaborations