This study explores a novel application of harmonically modulated chlorophyll fluorescence imaging for detecting phosphorus nutrient stress in plants. This approach may provide advantages over contemporary methods, appearing less susceptible to spurious light sources, requiring lower computational demand than hyperspectral imaging, and enabling faster results compared to the rapid light-induced fluorescence method, which requires extended dark adaptation. A sinusoidally modulated fluorescence imaging system was developed to capture fluorescence signals from wheat plants under varying phosphorus conditions. Signals were processed using wavelet scattering, wavelet packet, and fast Fourier transforms, then classified via support vector machine. Results indicate that the method can distinguish between phosphorus-sufficient (100%), phosphorus-deficient (50%), and phosphorus-deprived (0%) plants, achieving up to 98% accuracy in binary classifications under the tested condition between high and low phosphorus levels, and 79% –98% accuracy for intermediate comparisons (high versus medium, medium versus low). Wavelet packet transform consistently delivered the highest performance ($>$96% accuracy across all pairs) with minimal variability. The study establishes harmonically modulated chlorophyll fluorescence as a promising tool for rapid, noninvasive nutrient stress detection, advancing precision agriculture practices.