Crop Yield Recommendation System Using Machine Learning

Abstract

Machine learning techniques are revolutionizing multiple industries; various researches have been put forward as regards mitigating pest and disease effects on food production. The ability to identify plant diseases on time can help reduce the level of destruction caused by the diseases. This research focuses on tomato leaf disease symptoms which include leaf spots, blights, stem rots, plant wilt, Leaf necrosis, and leaf yellowing using Convolutional Neural Network (CNN) as a classification technique and also using keras and tensorflow Python machine learning libraries to build a model for plant disease classification. Convolutional Neural Networks (CNN) can automatically recognize interesting areas in images which reduces the need for image processing, and training images. The dataset was obtained from plantvillage.org and was split into training, testing, and validation sets, the training images were augmented and fed into a CNN model for training the model and were then tested on the test set to check against overfitting before finally using it to detect disease on the validation set which showed very positive results. The experiment conducted on the Tomato Leaf disease dataset comprising six (6) different classes achieved an overall accuracy of 98%. Hence the results from this research show that CNN can be used to develop highly efficient plant disease detection models.