Prediction of cut rose growth stages and leaf color variation using image analysis method

Predicting crop growth stages, especially harvest time, plays a very important role in planning greenhouse production. There are many studies on the application of digital image analysis technology to estimate crop growth behavior in greenhouses. In the present study, changes in leaf color characteristics of four commercial rose cultivars over time were investigated using color image processing using image j software and RGB color space. The results showed that there was a high correlation between leaf color components and stem growth stage in cultivars with white flowers (R2 = 0.986) and cultivars with colored flowers (R2 = 0.94), and a significant difference was also observed between color components in leaves of different stem layers. Also, a good correlation was obtained between direct measurement of total chlorophyll by spectrophotometry and chlorophyll index by SPAD. Among the models studied, it was found that the linear model and the exponential model performed better in establishing a logical relationship between data obtained from stem height and leaf color changes, although differences were observed between cultivars in this regard. The ability of the image analysis method to non-destructively detect color changes between leaf layers and establish a meaningful and logical link between stem growth changes was recognized as valuable and worthy of attention. Developing this model for other important greenhouse rose cultivars could provide greenhouse rose producers with a robust and reliable tool to help them adjust production schedules and predict harvest and market timing.