A Novel System to Evaluate Stereo Vision by Polarized Dual-Projection

Abstract

 

Purpose:

 

To present a novel system for evaluating human stereo vision by estimating disparity and perception time.

 

Methods:

 

The system uses polarized dual-projection in combination with polarized 3D glasses to provide 3D images with variable disparity. The displayed 3D scene is generated by our special software framework which we created to enable the development of arbitrary stereo vision patterns. In a first step stereo vision with respect to perception time was estimated by showing two disks with different disparities in varying sizes. The observer had to decide which disk is located closer to him. Depth differences could only be determined by the disparity of the disks and not by size or illumination differences. During the test the disk sizes were varied to five diameters (23940'';17990'';4711'';2943'';1137''). The detection time to identify the closer disk was stored in each step. For each size the same disparity differences were shown which ranged from 162’’ to 975’’. In a preliminary study five healthy subjects (25-28 years, 5 males) were measured.

 

Results:

 

The detection times show a direct relation to object sizes and vary from 1266 ms to 4438 ms. With decreasing disk sizes we noted increasing detection times for correct decisions (see diagram). All subjects show for each size level comparable detection times. Only the detection times of subject 5 differ significantly for decreasing disk sizes. All subjects correctly detected more than 50% of the closer disks in each size level.

 

Conclusions:

 

The presented test system provides the ability to generate arbitrary visual patterns to measure stereo acuity and detection times simultaneously. The system enables an implementation in regular 3D-TV-screens to evaluate human stereo vision.  

 

Keywords: binocular vision/stereopsis • depth • clinical (human) or epidemiologic studies: systems/equipment/techniques