Multimodal Visual Sensing of Temperature and Pressure: From Spectroscopic Readout to Multiple Linear Regression-Enhanced RGB Analysis

Abstract

Transforming a single luminescent host into distinct temperature- and pressure-sensing systems through composition control provides a powerful route toward multifunctional optical sensors. Here, we introduce a concentration-tunable KGaGeO4:Bi3+,Eu3+ platform whose sensing function can be selectively directed toward thermometry or manometry by adjusting the dopant balance. Spectroscopically distinct Bi3+ centers and Eu3+ emission exhibit differentiated responses to temperature and pressure, enabling multimodal readout through ratiometric luminescence, chromaticity coordinates, spectral shifts, and visible color changes. Importantly, these color changes were translated into quantitative temperature and pressure maps using RGB imaging combined with multiple linear regression (MLR). By simultaneously exploiting multiple color channels, MLR increased the maximum relative thermal sensitivity from approximately 1% K-1 for conventional RGB ratios to 8.8% K-1 and the pressure sensitivity from 169% GPa-1 to nearly 721% GPa-1. This work presents the first application of MLR-assisted RGB analysis for quantitative luminescence pressure sensing. The proposed strategy integrates composition-controlled functionality, multimodal spectroscopic sensing, direct visual readout, and data-assisted imaging, demonstrating the synergy between material engineering and multivariate analysis for highly sensitive multifunctional optical sensing.