Structural parameterisation and immersive validation for cross carrier representation of Mongolian patterns
Year: 2026
Type: article
Abstract: Mongolian patterns are routinely digitised as images, but their structural relations require explicit representation for comparison across material carriers. We developed a framework linking line direction, element boundaries, and compositional hierarchy to cross-carrier comparison and immersive behavioural validation. The study analysed 566 field-documented images from six carrier types and used a within-subject design with 21 non-specialist participants. Alignment-event rates were 70.5% for the full model, 29.1% without the direction field, and 43.2% without the element layer. All four prespecified participant-level comparisons remained significant after false-discovery-rate correction. The ablations revealed distinct contributions from structural layers. Patterns with recoverable dominant structural logic showed stronger cross-carrier consistency. Behavioural alignment is interpreted as an association between participant orientation and extracted structural relations. Within the tested corpus and architecture, the framework provides an operational approach to assessing structural continuity across carriers and after immersive remapping.
Source: npj Heritage Science
Authors: Jiatong Liu, Yue Huang
Institution: Beijing University of Posts and Telecommunications
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W7218479288
FWCI:
Cites:
Cited by:
Related to: