Funding entity Comunidad de Madrid
Traditional fumigation of plantations is applied with the same amount of chemicals to the whole surface. As a consequence it is very common to find devastated areas of crop, short in fumigation, and other areas where chemicals had probably been overused unnecessarily. Automatic fumigation ground vehicles are unable to detect which areas may require additional treatment, hence applying uniform fumigation. In general excessive chemicals are used, involving extra economical costs and important environmental impact. A more effective approach is based on the analysis of aerial photography to identify those regions requiring additional treatment and applying fumigation selectively. The only problem is that image acquisition and processing is usually very expensive and too slow to be fully effective. This project will develop an artificial vision system installed onboard a small semi-autonomous helicopter that will survey the plantation in different altitudes diagnosing crop damaged areas accurately in real-time. As result a geo-referred fumigation map is estimated, in order to help an autonomous land vehicle in charge to fumigate.