DRCmpVis: Visual Comparison of Physical Targets in Mobile Diminished and Mixed Reality

Area of Science:

• Computer Science
• Human-Computer Interaction
• Virtual Reality

Background:

• Physical objects are often presented in stereotyped ways, limiting comparison by multiple attributes.
• Existing methods for re-grouping or re-ranking objects require breaking original layouts or extensive manual effort.

Purpose of the Study:

• To propose a novel mobile framework, DRCmpVis, for in-context multi-attribute comparisons of physical objects.
• To enable intuitive, flexible, and seamless visual comparison strategies without altering physical object arrangements.

Main Methods:

• Utilizing mobile cameras to scan physical objects in the real world.
• Employing convolutional neural networks for object segmentation and labeling.
• Replacing scanned objects with virtual avatars in a digital reality (DR) environment for comparison.
• Implementing filtering, re-grouping, and re-ranking strategies on these virtual avatars.

Main Results:

• DRCmpVis facilitates efficient object searching and multi-attribute visual comparisons by leveraging virtual space computation.
• Demonstrated usability, expressiveness, and efficiency through user studies, assessments, and case studies.
• The framework successfully avoids disrupting original physical object layouts during comparison.

Conclusions:

• DRCmpVis offers a novel approach to enhance object comparison by integrating real-world scanning with virtual manipulation.
• The proposed framework provides a flexible and efficient solution for multi-attribute comparisons in various scenarios.
• The study validates the practical application and effectiveness of DRCmpVis in improving object interaction and analysis.