Detection and Classification of Leaf Disease using Machine Learning Approach

  • Arpana Mahajan et al.
Keywords: Histogram Equalization, Gaussian Filter, Haralick, Random forest


We propose and tentatively assess a product answer for programmed recognition and characterization of plant leaf diseases. Moreover, even though that person has always been bad in biology but would like to know more about that plant. It simply means that he/she is interested to explore his/her knowledge in this particular area. He might be interested to know its name or about its specific features. Sometimes, he/she might be interested to search a plant if it is rare or on the verge of the existence. Indeed, even today, distinguishing proof and arrangement of obscure plant species are performed physically by master individual who are not many in numbers. Here, we are introducing another acknowledgment approach dependent on Leaf Features Fusion and Random forest (RF) Classification calculations for characterizing the various kinds of plants. The proposed methodology comprises of four stages that are preprocessing, division, highlight extraction and grouping stages. Since most sort of plants have one of a kind leaves. Leaves are not exactly equivalent to each other by attributes, for example, the shape, shading, surface and vein. There are numerous highlights of leaf, for example, Color highlights, Vein highlights, GLCM highlights, Shape highlights and Gabor highlights. These all highlights are melded by idea linking of two vectors. Along these lines, order approach displayed in this exploration relies upon plant leaves. Test results show precision and different parameters estimated in this methodology with combination of every one of these highlights or their various mix. This is a clever framework which can recognize tree species from photos of their leaves and it gives precise outcomes in less time. Here we are developing and C# base Web application that support small farmers in analyzing the damage to their plants quickly and easily.