Development of an Intelligent System for Diagnosis of the Botrytis Elliptica Disease in the Lilium Plant Using Image Processing

The automatic detection of plant diseases in the early stages of growth can increase the quality of the final product and prevent the occurrence of permanent damage in large part of farms. Therefore, in this research an intelligent system was designed and developed based on image processing in order to detect and eliminate the disease in the lilium plant leaf, as well as the classification of healthy plants from the unhealthy ones. Accordingly, 20 healthy flowers and 20 unhealthy were evaluated by machine vision system. In order to classify plants, 19 color and morphology parameters of the plant were extracted and the most effective ones (leaf L, leaf a, leaf b, stem L, and stem length) were selected by fuzzy entropy method and these suitable features were grouped by the similarity classifier. As result, the efficiency of the proposed algorithm to diagnose and classify the disease using fuzzy entropy H1, H2 / H3 fuzzy entropy and without applying selection of features method were 96.15, 93.18 and 84.3, respectively.