Advanced Machine Learning Algorithm Based System for Crops Leaf Diseases Recognition

Crops diseases can create a major economic loss for a state-run. Overcoming on that issue is the main requirement. In this work, we propose an automated system for recognition of potato and corn leaf diseases. Three core phases architecture includes handcrafted features are extracted such as histogram-oriented gradient (HOG), Segmented Fractal Texture Analysis (SFTA) and local ternary patterns (LTP). In the second phase, principal component analysis (PCA) along entropy Skewness based score values are computed and resolve the problem of curse of dimensionality. In the last phase, classification is performed using various classifiers. The Plant Village dataset is utilized for validation and classify selected potato and corn diseases. Competent results are obtained in the range of 92.8% to 98.7% on chosen crops diseases which are better as compare to existing techniques.