REMOTE SENSING OF CUCUMBER POWDERY MILDEW USING ADVANCED UNMANNED VEHICLE AND IMAGE PROCESSING TECHNIQUES

Pesticides is considered one of the most hazardous compounds that caused serious health implications to man and environment. However, their use is perquisite to guarantee a high yield especially, for greenhouse cucumber production. In fact, the process success depend on spraying coverage, rate and early detection of infection symptoms. In this study, a novel remote sensing approach for detecting Powdery Mildew symptoms in cucumber (Cucumis sativus) leaves has been developed. The experiment was conducted under greenhouse conditions. An unmanned vehicle platform used to carry the remote sensing sensors. Infected leaves were then analyzed using a pioneer image processing model. In this model, leaves were classified according to different feature especially, the reflect color. Image processing results were then assessed using ground reference data (lab analysis). The accuracy assessment of remotely-sensed data was ground reference data were comparable. Because changes in infected leaves associated with the imagery model data, we conclude that remotely-sensed data derived from the unmanned vehicle sensor data hold promise for detecting Powdery Mildew symptoms in cucumber leaves.