Plants have developed an imperative source of energy and aimportant problem in solving the problem of global
warming. However, plant diseases threaten the livelihoods of this important source. Convolutional Neural Networks (CNN) have
shown excellent performance (more than humans) in recognizing problems and image classification problems. This paper
describes the possibility of classifying plant diseases with leaf images taken by CNN in the natural situation. The model is
designed based on the Le-Net architecture for organization of soybean plant diseases, including healthy leaf images taken from
the plant village database. Performance evaluations on several popular benchmark datasets prove that our method is better than
the latest technology. This article proposes a method to improve the classification performance of classes with few training
examples
