This study proposes a design-driven framework
for restoring damaged museum artifacts using artificial intelligence (AI) and
augmented reality (AR). Based on an evaluation of current virtual restoration
methods, seven conceptual scenarios were formulated, with Scenario 6
implemented as a functional AR prototype.
Three
primary research questions guided this study:
RQ1:
How can AI and AR be combined to restore incomplete artifacts while maintaining
historical authenticity and enhancing public engagement?
RQ2:
Which restoration scenario provides the most efficient, scalable, and accurate
outcome for museum workflows?
RQ3:
How can AR presentation balance interpretive depth with curatorial integrity?
Two
hypotheses were evaluated:
H1:
Integrating AI inpainting with AI-driven 3D modeling significantly reduces
restoration time compared to manual or hybrid workflows.
H2: Presenting AI-restored artifacts via AR
improves visitor engagement and contextual comprehension compared to static
displays.
The results supported H1 and provided
preliminary validation for H2, demonstrating the practicality of combining AI
and AR for digital artifact reconstruction and museum display.
