Methods of Health Improving Using Leaf Image Processing

The phytopathological condition of the plant can all be detected in the leaf by the appearance of visible symptoms in the form of spots, discoloration. Deep learning techniques develop solution in various computer vision problems, which inspires us to use them to vine diseases identification task. These methods and systems to prevent plant diseases are very topical in order to reduce the work of farmers. This paper presents methods to identify vine diseases by species using leaves photos. Deep learning methods are designed in images processing to identify spots on leaves. The shape of the leaf is useful for identifying the vine species. After the images processing results are compared the methods with the performance are saved for the next analysis. The processing is performed in the Tensorflow programming environment in order to implement an automatic vine analysis system.