LEAF INCLINATION BASED NON DESTRUCTIVE WATER STRESS INDICATION FOR VEGETABLES

In the present work a machine vision system is presented for analysing side projected canopy images of tomato plants. The images were taken by a digital camera in an experimental model greenhouse. The camera acquired images at user defined time-scales. The software measured the growth of the plants in vivo and in situ and focused on the leaf and stem edge inclination states under quasi-optimal and stress conditions. Stems were defined as all of the branches and petioles on the plant. ‘Leaf and stem’ inclination was calculated from the angle of the edge lines of the whole of the visible part of the plant canopy, including leaves, petioles and branches. Around each edge point there are a few pixel-long edge lines. The direction of these lines, located usually on the outside of the stems and leaves, are compared to the horizon. The general inclination value of the monitored canopy is counted from the direction of these lines in the range 0 to -90 degrees. This inclination value can provide non destructive information about plants’ wellness or stress condition. Besides irrigation the information thus gained can be used for setting optimal greenhouse parameters and measure plant growth.