Use of Computer Vision Methods for Irrigation Control

Monitoring of greenhouse plants gives a straight diagnosis on their health and wellness. A computer vision method was investigated for detection and localization of leaves’ wilting as a sign of visual appearance. Tomato, cucumber and paprika were tested using side projected canopy images. The camera was focused on the plant canopy from horizontal direction in order to obtain images covering the leaves. At defined time scales images were taken to analyze leaf and stem movements as the visual signs of plant state. The machine based control can be applied successfully in crop irrigation. The parameters of the control unit were identified and adjusted in each case of the plant, as well.