ABSTRACT

A variety of diseases attack rice, one of India’s most widely planted crops, at various phases in the growing process. It is exceedingly difficult for farmers with poor understanding to identify these diseases manually. Automated picture identification systems based on Convolutional Neural Network (CNN) models are showing considerable promise in deep learning research recently. Our deep learning model was created using Transfer Learning on a small dataset because it was difficult to find an image dataset of rice leaf disease. VGG-16 is utilized to train and assess the proposed CNN architecture, which is based on rice field and internet datasets. The proposed model has a 95 percent accuracy rate. Deep Learning, Convolutional Neural Network (CNN), fine-tuning and rice leaf diseases are some of the terms in this index.

EXISTING SYSTEM:

In the past, the only way to identify a disease was to manually examine the leaf. The illness was discovered using a combination of visual inspection of plant leaves and consulting a reference book. This method has three key drawbacks: limited accuracy, the inability to analyse every leaf, and a lengthy time investment. As science and technology progress, new methods for accurately diagnosing these diseases emerge. Image processing and deep learning are two methods to consider. Filtering, clustering and histogram analysis are some of the approaches used in image processing to identify the diseased area. Deep learning neural networks, on the other hand, are used to detect disorders.

3.1.1 DISADVANTAGES OF EXISTING SYSTEM:

• It’s impossible to study every leaf, and it takes a long time.

• PROPOSED SYSTEM:

A VGG16 transfer learning neural network is used in this paper to train a dataset of rice diseases, and the trained model may be used to predict disease from new photos. Because the Rice Leaf dataset from KAGGLE was too tiny for author to train the VGG16 model, he turned to the transfer learning CNN algorithm, which transfers an existing CNN model to a new dataset and then uses the new data to train the model.

It has been shown that VGG16 transfer learning improves prediction accuracy in both a normal CNN model and a normal CNN model with VGG16 transfer learning.

3.2.1 ADVANTAGES OF PROPOSED SYSTEM:

• • VGG16 can be utilised to identify rice leaf diseases. Over ninety-five percent of the time, it has been right.

SYSTEM REQUIREMENTS
SOFTWARE REQUIREMENTS:
• Programming Language : Python
• Font End Technologies : TKInter/Web(HTML,CSS,JS)
• IDE : Jupyter/Spyder/VS Code
• Operating System : Windows 08/10

HARDWARE REQUIREMENTS:

 Processor : Core I3
 RAM Capacity : 2 GB
 Hard Disk : 250 GB
 Monitor : 15″ Color
 Mouse : 2 or 3 Button Mouse
 Key Board : Windows 08/10