Detecting fluorescence using consumer-accessible imaging devices could enable low-cost authentication and sensing technologies. Here we present a smartphone-based fluorescence detection method that exploits decorrelation between \(RGB_{\textrm{det}}\) camera channels under controlled illumination. Using 16-bit smartphone images, we measured responses from fluorescent and colour-matched non-fluorescent samples under 125 illumination conditions defined by combinations of red, green, and blue LED intensities. Pearson correlation coefficient (PCC) values were calculated across the pixel intensities within cropped image regions, quantifying the statistical relationship between pairs of camera colour channels, (\(R_{\textrm{det}}\), \(G_{\textrm{det}}\), \(B_{\textrm{det}}\)), to assess inter-channel relationships. Non-fluorescent samples exhibited consistently strong correlations (PCC \(\approx\) 1), reflecting proportional scaling of reflected light across channels. In contrast, fluorescent materials produced broader PCC distributions spanning − 1 to 1 due to Stokes-shifted emission that disrupts linear channel relationships. Simple PCC thresholding enabled robust discrimination between fluorescent and non-fluorescent samples across illumination conditions. This RGB-PCC approach demonstrates that fluorescence signatures can be detected using correlation structure rather than absolute intensity, enabling a low-cost, hardware-efficient method for visible-light fluorescence detection with potential applications in authentication, materials analysis, and portable sensing.
