Monitoring the health of vineyard using drone

Machine learning techniques are used for image processing to identify the presence of a grapevine disease at an early stage. The drone images are pre-processed before being used in the disease detection software so that only the area of interest, i.e. the leaf, can be analysed, eliminating details that could cause errors in the image analysis process. Following the analysis, the user receives a graphical response in the same software platform.

In order for the software using Machine Learning techniques to be able to detect diseases on the vine, it is necessary to upload the image from the drone to be processed. The next step is to pre-process the image where areas of interest are selected. To make disease detection as accurate as possible, the image is segmented to leave only the area of interest. The last aspect of the detection process is the analysis of the area of interest by the software developed using Machine Learning techniques.