Measuring Depth Perception for Stereoscopic Images Using a Phase-shift Model

We present an effective depth perception assessment index for stereoscopic images using a phase-shift model. To be more specific, we use Gabor filter to compute the responses of left and right images respectively, and proposed a phase-shift model for computing disparity maps based on phase gradient and phase difference information. Then, quality score for depth perception is obtained by measuring the similarity between the estimated disparities of the original and distorted stereoscopic images. Experimental results on two publicly 3D image quality assessment databases demonstrate that, in comparison with the most related existing methods, the devised algorithm achieves high consistency alignment with subjective assessment.