QoE-Guided Warping for Stereoscopic Image Retargeting

Area of Science:

• Computer Vision
• Human-Computer Interaction
• 3D Display Technology

Background:

• Stereoscopic 3D (S3D) image retargeting is crucial for adapting content to various display resolutions.
• Existing methods often neglect the user's Quality of Experience (QoE) during retargeting.
• Optimizing visual comfort and depth perception is vital for effective S3D viewing.

Purpose of the Study:

• To develop a novel QoE-guided warping method for stereoscopic image retargeting.
• To enhance user's QoE by considering image quality, visual comfort, and depth perception.
• To adapt stereoscopic images to target resolutions while preserving key visual attributes.

Main Methods:

• A QoE-guided warping approach is proposed for stereoscopic image retargeting.
• The method simultaneously optimizes 2D coordinates and 3D depth information.
• Energies for shape preservation, visual comfort, and depth perception are incorporated, considering viewing configurations.

Main Results:

• The proposed method demonstrates improved performance across QoE factors compared to existing techniques.
• Experimental results show optimization of image quality, visual comfort, and depth perception.
• The approach leads to a more promising overall S3D experience.

Conclusions:

• The QoE-guided warping method effectively retargets stereoscopic images while enhancing user experience.
• Integrating visual comfort and depth perception significantly improves S3D content adaptation.
• This approach offers a valuable contribution to the field of stereoscopic image processing and display.