Consumer Depth Cameras for Computer Vision: Research Topics by Jan Smisek, Michal Jancosek, Tomas Pajdla (auth.), Andrea

By Jan Smisek, Michal Jancosek, Tomas Pajdla (auth.), Andrea Fossati, Juergen Gall, Helmut Grabner, Xiaofeng Ren, Kurt Konolige (eds.)

The capability of customer intensity cameras extends way past leisure and gaming, to real-world advertisement functions. This authoritative textual content studies the scope and effect of this quickly growing to be box, describing the main promising Kinect-based examine actions, discussing major present demanding situations, and showcasing intriguing purposes. good points: offers contributions from a world choice of preeminent specialists of their fields, from either educational and company learn; addresses the vintage challenge of multi-view geometry of ways to correlate photographs from diversified viewpoints to concurrently estimate digital camera poses and international issues; examines human pose estimation utilizing video-rate intensity pictures for gaming, movement seize, 3D human physique scans, and hand pose popularity for signal language parsing; offers a evaluation of methods to numerous attractiveness difficulties, together with type and example studying of items, and human job reputation; with a Foreword by way of Dr. Jamie Shotton.

Show description

Read Online or Download Consumer Depth Cameras for Computer Vision: Research Topics and Applications PDF

Similar research books

The Easy Guide to Repertory Grids

A common advent to the robust psychological mapping device of repertory grid method. Repertory grid approach is a method for picking out, intimately, what you or a person else relatively thinks approximately a subject. you should use it as a device for private discovery, as a tool for workforce development actions, or as a problem-solving relief.

Operations Research Proceedings 2010: Selected Papers of the Annual International Conference of the German Operations Research Society

This booklet comprises chosen papers from the symposium "Operations study 2010" which used to be held from September 1-3, 2010 on the "Universität der Bundeswehr München", Germany. The overseas convention, which additionally serves because the annual assembly of the German Operations learn Society (GOR), attracted greater than six hundred individuals from greater than thirty international locations.

Genetic Research in Psychiatry: “2. Münchner Genetikgespräche” September 12 – 15, 1991 C.I.N.P. President’s Workshop

Significant and fascinating advances in psychiatric genetics have been mentioned at a 3-day foreign workshop. The across the world popular editors assembledan outstanding record of experts, all of whom are major of their topic. stories and brief articles which pressure designated difficulties or new examine effects were introduced jointly during this e-book, additionally together with extensive discussions of the various themes.

Meanings and Motivation in Education Research

Meanings and Motivation in schooling learn demonstrates the colourful and very important connection among the researcher and their study. examine is frequently perceived as an entity that's possible produced in isolation; although, there are lots of very important components that are concerned, and this booklet explores the complexities of investigating the categorical components of which means and motivation for researchers operating within the schooling quarter.

Extra info for Consumer Depth Cameras for Computer Vision: Research Topics and Applications

Example text

Pattern Anal. Mach. Intell. : SLAM combining ToF and high-resolution cameras. In: IEEE Workshop on Applications of Computer Vision, pp. : A nearest neighbor data structure for graphics hardware. : Accelerating nearest neighbor search on manycore systems. : Object modelling by registration of multiple range images. Image Vis. Comput. : A volumetric method for building complex models from range images. In: Conference on Computer Graphics and Interactive Techniques, SIGGRAPH, pp. 303–312. : Color constrained ICP for registration of large unstructured 3D color data sets.

For the nearest neighbor search, being a major bottleneck in terms of runtime, CPU architectures have shown to benefit from space-partitioning data structures like k-d trees [1]. In contrast to algorithmic improvements, hardware acceleration techniques are increasingly attracting the attention of the community. Garcia et al. have shown that a GPU-based brute-force implementation outperforms a CPU-based k-d tree [15]. The reason for this lies in the fact that the brute-force primitive can be implemented efficiently using techniques known from the well understood problem of GPU-based matrix–matrix multiplication.

Compared to a conventional ICP that relies on the pure 3-D geometry [4, 9], incorporating photometric appearance as a complementary source of information is advantageous in cases of non-salient surface topology, recall Fig. 1 and the experimental results in related work [24]. Approaches that combine dense geometric point associations with a sparse set of correspondences derived from local photometric features are limited to interactive frame-rates, as feature extraction is computationally expensive even if performed on the GPU [19].

Download PDF sample

Rated 4.56 of 5 – based on 22 votes