In this paper we are considering the great outcome of Convolution Neural Networks (CNNs) in the field of images. Here we utilize the convolution systems to show improvement in the model to identify the plant sickness. An innovative method inspires a simple framework which classifies the sickness from sound leaves in speedy manner. In the model which will be discussed will apparently analyze in identification of nearness of leaf followed by the sound recognition of leaves and maladies in 4 distinct. It has the capacity to recognize plant leaves form their environment. Every basic requirement to implement this model has completely represented through the paper, right from getting pictures to build the database till survey by farming specialist. The experimental results of this malady acknowledge model shows the accuracy in between 92% to 98% , for isolated class test at the maximum of 96.3%