312 Figure 5. Statistical comparison of positioning error under different correction strategies (no correction, IMU correction, and cumulative position correction) for solo GNSS positioning and RTKGNSS. Box plots indicate median, interquartile range, and distribution of positional error. Table 3. Quantitative positioning accuracy obtained under different correction strategies for solo GNSS positioning and RTK-GNSS. No correction IMU Correction Cumulative position correction Solo positioning 1.54m 3.35m 2.00m RTK-GNSS 0.83m 2.67m 0.46m Latency between GNSS acquisition and AR rendering remained within limits acceptable for walking-speed bench inspection, supporting stable visualization without perceptible delay. These characteristics indicate that the proposed system is well suited to short-range grade verification, supports grade-control interpretation, enhances situational awareness, enables visual verification of ore-waste boundaries. Overall, the results demonstrate that polygon-based ore modeling combined with RTK-GNSSanchored augmented reality can deliver spatially coherent, operationally meaningful grade visualization at the bench scale. This establishes a technical foundation for future quantitative validation, extended field deployment, and integration with real-time digital-twin mining workflows. 4. CONCLUSION This study introduced a smart-glass-based augmented-reality framework integrated with RTKGNSS positioning for real-time bench-level ore-grade visualization in open-pit mining. By converting conventional block-model grades into lightweight polygon representations and synchronizing them with
RkJQdWJsaXNoZXIy MTM0Mzk2