Analisis Korelasi Warna Daun Tomat dengan Kandungan Klorofil Menggunakan Pencitraan RGB dan Spektrofotometri

Abstract

This research investigated the relationship between leaf color (RGB values) and chlorophyll content in tomato plants as a non-destructive method for chlorophyll estimation, offering a potential alternative to spectrophotometry. The study employed an experimental design, analyzing 300 leaf samples categorized by age (young, middle, old) from tomato plants. Leaf images were acquired using a scanner in RGB mode, and chlorophyll content was measured using spectrophotometry after acetone extraction. Data analysis involved fitting linear regression models (Model 4.1: using R, G, B values; Model 4.2: using ratios of R, G, and B to reduce lighting effects) to predict chlorophyll content (chlorophyll a, chlorophyll b, and total chlorophyll) from RGB data. Results showed a significant positive correlation between green intensity (G) and chlorophyll content, particularly chlorophyll a. Model 4.2, incorporating relative color ratios, demonstrated improved accuracy in predicting chlorophyll a (0.8583) and chlorophyll b (0.7545) compared to Model 4.1 (0.8576 and 0.7522, respectively). While the RGB imaging method proved efficient and relatively accurate, particularly for chlorophyll a, the accuracy for chlorophyll b and total chlorophyll remained lower. The study concludes that RGB imaging, especially with the modified Model 4.2, offers a promising non-destructive alternative for chlorophyll estimation in tomatoes, but further research using advanced machine learning techniques and broader environmental conditions is recommended to enhance accuracy and generalizability.