Runtime Analysis of GPU-Based Stereo Matching

Abstract

This paper elaborates on the possibility to leverage the highly parallel nature of GPUs to implement more efficient stereo matching algorithms. Different algorithms have been implemented and compared on the CPU and the GPU in order to show the speedup gained by moving the computation to the graphics card. The results were evaluated for accuracy using the test available on the Middlebury website for stereo vision. An assessment of the runtime performance was done by a script which examined the runtime behaviour of the individual steps of the stereo matching algorithm.

Authors and Affiliations

Christian Zentner, Yan Liu

Keywords

Related Articles

Gatekeepers Practices in Knowledge Diffusion within Saudi Organizations: KFMC Case Study

Gatekeepers in organizations play a critical role in terms of disseminating and transferring outside knowledge into their groups. This research contributes in identifying the gatekeepers' practices in terms of gathering,...

Correlated Topic Model for Web Services Ranking

With the increasing number of published Web services providing similar functionalities, it’s very tedious for a service consumer to make decision to select the appropriate one according to her/his needs. In this paper, w...

LINK-BUDGET DESIGN AND ANALYSIS SHOWING IMPULSE-BASED UWB PERFORMANCE TRADE-OFF FLEXIBILITY AS INTEGRATOR SOLUTION FOR DIFFERENT WIRELESS SHORT-RANGE INFRASTRUCTURES

Future wireless indoor scenarios are expected to be complex requiring wireless nodes to adaptive responding to dynamic changes according to channel conditions. Interacting with neighboring nodes to achieve optimized perf...

Expensive Optimisation: A Metaheuristics Perspective

Stochastic, iterative search methods such as Evolutionary Algorithms (EAs) are proven to be efficient optimizers. However, they require evaluation of the candidate solutions which may be prohibitively expensive in many r...

Securely Eradicating Cellular Dependency for E-Banking Applications

Numerous applications are available on the Internet for the exchange of personal information and money. All these applications need to authenticate the users to confirm their legitimacy. Currently, the most commonly empl...

Download PDF file
  • EP ID EP106588
  • DOI 10.14569/IJACSA.2015.061138
  • Views 103
  • Downloads 0

How To Cite

Christian Zentner, Yan Liu (2015). Runtime Analysis of GPU-Based Stereo Matching. International Journal of Advanced Computer Science & Applications, 6(11), 287-291. https://europub.co.uk/articles/-A-106588