Response of sunflower to irrigation, boron spraying and plant density under new valley conditions [electronic resource].

Includes reference.

This research aimed at overcoming some of the limitations encountered in the COUISe of plant leaf shape analysis studies. Mathematical morphology of plant leaf venatiou was exploited as a shape analysis technique for plant identification purposes. Leaf shape features were extracted from preprocessed leaf color images using a pre-selected set of twenty structuring elements to generate a distinctive leaf signature for individual plants. Prototype leaf signatures were produced from a training set for three plants, namely: Alfalfa (Medicago Sativa), Mint (Mentha Piperita), and Pomegranate (Punica GranatumI) to calculate decision functions for developing a Minimum Distance Classifier. Four factors that could influence the performance of the classifier were considered in forming a testing set of 162 leaf color images. The four factors considered were: plant type, leaf orientation angle, leaf size, aod leaf occlusion ratio. The classifier effectively handled the testing set by correctly designating leaves to the plants to which they belong. A perfect performance was achieved in identifying the three plants with correct classification rates of 100%, 96.3%, aod 100% for Alfalfa, Mint, and Pomegranate, respectively. Key words: Precision agriculture, Shape analysis, Leaf venation, Morphological features, Template matching, Leaf signature, Plant identification.

Summary in Arabic.

1