Multi-channel Image Analysis for Monitoring of Sodium Fluorescein Concentration in Clays

To quantify sodium fluorescein concentrations in clays, lab-scale diffusion experiments were conducted using acrylic flow chambers filled with glass beads and three different types of clay (kaolinite, montmorillonite, and bentonite). Images of the clay layer were captured, split into RGB channels, and then transformed to grey scale to measure color intensities. A one-dimensional diffusion model was used to compare the simulated and measured concentration data for evaluating the feasibility of the multi-channel image analysis technique. For kaolinite and montmorillonite, the measured and simulated data showed good agreement during the entire experiment using blue (E = 0.95) and green (E = 0.96) channels respectively. This correspondence resulted from a gradual decrease in fluorescein concentrations with an increase in color intensities of blue and green channels for kaolinite and montmorillonite. For bentonite, a favorable agreement between the measured and simulated concentrations was shown during the early half of loading time (8.2 PVs) using only the blue channel (E = 0.98), while there was a great discrepancy between the measured and simulated data after 18.1 PVs. The multi-channel image analysis combined with blue and green channels for bentonite was employed to determine the fluorescein concentration and showed a better fit (E = 0.96) compared with only one channel (E = 0.56 for the blue channel and E = 0.42 for the green channel). As a result, multi-channel image analysis could be used where there was no variance in color intensities of one channel despite an increase in fluorescein concentration in clay.