MULTI-CLASS CLASSIFICATION OF PLANT LEAF DISEASES USING FEATURE FUSION OF DEEP CONVOLUTIONAL NEURAL NETWORK AND LOCAL BINARY PATTERN

Abstract

Crop diseases pose a significant risk to food safety, but rapid identification of evidence remains difficult due to a lack of critical infrastructure in many parts of the world. The result of precise methods in the field of pagebased image distribution is very well demonstrated. This paper uses random forest to identify healthy and diseased leaves from the generated data. Our implementation includes several levels of implementation, such as dataset creation, feature extraction, training classifiers, and classification. Data from diseased and healthy leaves were collected in a random forest to classify diseased and healthy images. We use Histogram of Oriented Gradients (HOG) to extract shape from images. Overall, using machine learning to train on large publicly available datasets allows us to clearly see the presence of diseases in plants at scale.