Using digital image analysis to determine plant characteristics (color, shape, area, disease, etc.) could be a good solution for growers to increase the efficiency of greenhouses and reduce greenhouse labor cost.

The experiment was conducted in a research greenhouse located at the University of Tabriz in Iran. Four greenhouse cut rose cultivars namely, ‘Caribia’, ‘Full House’, ‘Cherry Brandy’ and ‘Polar Star’ were bought from a commercial greenhouse and then were transferred to the research greenhouse.

Images of roses were taken with a digital camera during a certain time of the day and in specific light conditions. The images were then introduced to the Image software in order to undergo some preparation processes. Image pre-processing, segmentation and background deletion were done.

After preliminary image processing, pixel values of each channel (Red, Green, and Blue) were extracted from the images and then normalized or average red (R), green (G) and blue (B) values of the RGB color model were calculated.

Results revealed acceptable correlations between the leaves color components and the stage of stem growth in all the cultivars, even in white cultivar, ‘Polar Star.’ Among the models fitted to the data, linear and exponential models appeared to be best fitted, both showing high positive relationship between the changes of leaf color and stem elongation over time.

This study was conducted by a horticulture postgraduate student, Ms. Sepideh Tahmasebi, who ran the study to determine the correlation between leaf color variations and the stages of stem growth in hybrid roses.

She conducted her research with her lecturer, Mr. Mansour Matloobi. The study revealed a higher correlation between the leaf color variations and the stages of stem growth in both white color and colorful cultivars.

Tahmasebi completed her postgraduate in Floraculture in 2014 and is now working to get ready for doctorate study at the Department of Horticultural Sciences, Faculty of Agriculture, University of Tabriz in Iran.