PT Journal
AU Pourazar, B
   Meruvia-Pastor, O
TI A Comprehensive Framework for Evaluation of Stereo Correspondence Solutions in Immersive Augmented and Virtual Realities
SO Journal of Virtual Reality and Broadcasting
PY 2017
VL 13(2016)
IS 2
DI 10.20385/1860-2037/13.2016.2
DE Augmented Reality; Binocular Stereo; Depth Sensing Cameras; Disparity; Display Resolution; Field of View; Head-Mounted Displays; Human Visual System; Stereo Correspondence; Stereoacuity; Virtual Reality
AB In this article, a comprehensive approach for the evaluation of hardware and software solutions to support stereo vision and depth-dependent interactions based on the specific requirements of the human visual system within the context of augmented reality applications is presented. To evaluate stereo correspondence solutions in software, we present an evaluation model that integrates existing metrics of stereo correspondence algorithms with additional metrics that consider human factors that are relevant in the context of outdoor augmented reality systems. Our model provides modified metrics of stereoacuity, average outliers, disparity error, and processing time. These metrics have been modified to provide more relevant information with respect to the   target application. We illustrate how this model can be used to evaluate two stereo correspondence methods: the OpenCV implementation of the semi-global block matching, also known as SGBM, which is a modified version of the semi-global matching by Hirschmuller; and ADCensusB, our implementation of ADCensus, by Mei et al.. To test these methods, we use a sample of fifty-two image pairs selected from the Kitti stereo dataset, which depicts many situations typical of outdoor scenery. Further on, we present an analysis of the effect and the trade-off of the post processing steps in the stereo algorithms between the accuracy of the results and performance. Experimental results show that our proposed model can provide a more detailed evaluation of both algorithms. To evaluate the hardware solutions, we use the characteristics of the human visual system as a baseline to characterize the state-of-the-art in equipment designed to support interactions within immersive augmented and virtual reality systems. The analysis suggests that current hardware developments have not yet reached the point where their characteristics adequately match the capabilities of the human visual system and serves as a reference point as to what are the desirable characteristics of such systems.
ER