A Low-cost Image Processing Based Technique to Estimate Chlorophyll in Winter Wheat

This paper presents a fast low-cost image processing method for the prediction of chlorophyll content in BARI gom, a winter wheat variety in Bangladesh. SPAD-502 was used as chlorophyll meter in this work and pictures of wheat leaf were taken by Smart Phone Camera with small adjustments. It is found that in the late vegetative stage, the “ $(\mathbf{a}^{\ast}-\mathbf{b}^{\ast})$ ” index from $\mathbf{L}^{\ast}\mathbf{a}^{\ast}\mathbf{b}^{\ast}$ color model has the most significant correlation with SPAD-502 chlorophyll data compared to other indexes from $\mathbf{L}^{\ast}\mathbf{a}^{\star}\mathbf{b}^{\ast}$ and HSV color model. Moreover, an equation is developed to calculate the chlorophyll content from “ $(\mathbf{a}^{\ast}-\mathbf{b}^{\ast})$ ” color index using linear regression method. Thereafter, the equation is tested against 36 samples in a random manner and an average accuracy of almost 90% is found over the range from 30 to 50.