Evaluation of Different Color Space Models for Estimation of Nitrate and Phosphate in Soil and Water Using a Smartphone-Integrated Imaging Device

To regulate agricultural inputs and monitor environmental pollution, this study offers a smartphone-integrated imaging device to quantify nitrate $\left( {{\text{N}}{{\text{O}}_3}^ - } \right)$ and phosphate $\left( {{\text{P}}{{\text{O}}_4}^{3 - }} \right)$ in soil and water samples. Using a smartphone’s digital camera, this cheap and portable device takes an image of the extracted solution being tested. The HSV color space model’s Value (V) component of the images was used to estimate the amounts of ${\text{N}}{{\text{O}}_3}^ -$ and ${\text{P}}{{\text{O}}_4}^{3 - }$ in soil and water samples. The limit of detection of the device for soil ${\text{N}}{{\text{O}}_3}^ -$, water ${\text{N}}{{\text{O}}_3}^ -$, soil phosphate, and water phosphate were calculated as 0.1 mg L ⁻¹ , 0.07 mg L ⁻¹ , 0.001 mg L ⁻¹ , and 0.02 mg L ⁻¹ , respectively, with good accuracy (≤1% bias) and precision (~2% residual standard deviation). Furthermore, to offer comparable results for the estimation of ${\text{N}}{{\text{O}}_3}^ -$ and ${\text{P}}{{\text{O}}_4}^{3 - }$ in soil and water, all the indices of the RGB, CMYK, and CIELAB color space models were examined and compared with the V-model. This cost-effective and user-friendly device provides reliable measurements comparable to a regular spectrophotometer.