A Sophisticated Deep Convolutional Neural Network for Multiple Classification of Apple Leaf Diseases

Abstract

The identification of crop diseases that is automated is critical because it simplifies the labor-intensive chore of supervising large farms while also identifying illnesses in the beginning stages to minimize plant degradation. Aside from the detrimental impact on plant health, this circumstance has a huge influence on a country's economy owing to decreased productivity. The current method of disease detection used by specialists is slow and unsuitable for big farms. A combination of an 11-layer CNN is used in our proposed technique. Its goal is to divide apple tree leaves into four groups based on images: healthy, scab, cedar rust, and black rot illnesses. On the training dataset, our suggested model attained an accuracy of 96.25%. In addition, the model has a 96% accuracy in recognizing leaves with various illnesses.