Nitrogen determination on tomato (Lycopersicon esculentum Mill.) seedlings by color image analysis (RGB)

摘要

In order to investigate the effectiveness of a new method based on color image analysis and the Minolta SPAD-502 chlorophyll meter for the diagnosis of nitrogen deficiencies of tomato seedlings, a field experiment was conducted. In this study, five levels of nitrogen fertilization were established so as to induce nitrogen deficiencies in tomato seedlings. Thirty-five days after sowing, total nitrogen was evaluated by laboratory analysis. The chlorophyll index was determined using a SPAD-502 chlorophyll meter. Also, color images were taken with a digital camera; the color images were processed in MATLAB in order to determine the averages of the red color, green color and the blue color. The relationships between variables were analyzed by linear regressions and a one way analysis of variance (p < 0.01). Results showed that color image analysis correlated better with the status of plant nitrogen than the SPAD. From the color image analysis, the red and blue colors were more accurate predictors of nitrogen status on plants with R2?above 0.89. Color image analysis provides an accurate and quick way for nitrogen estimation and can contribute for early detection of nitrogen deficiency in tomato seedlings. The SPAD method is not a reliable way to estimate the nitrogen status on tomato seedlings.