In recent years, data mining based image classification models have shown excellent performance in various applications. In the medical field, image classification can be used for the early diagnosis of diseases. We propose a framework using data mining algorithms and deep learning techniques for predicting the diseases in the plants. The system consists of two parts: (1) image data extraction, and (2) disease prediction. For the image data extraction part, we propose an improved CNN-based data extraction algorithm. From a huge number of images, the system can automatically extract pertinent visual data. We employ a CNN-based model to predict the disease from the retrieved picture data for the disease prediction portion. A dataset of photos representing plant diseases was used to assess the proposed framework. The results of the experiment demonstrated that the suggested system is capable of correctly predicting the disease from the visual data.

Disease Detection, Data Mining Deep learning, Tensorflow