Plant diseases identification and detection

What is it about?

The recent advancement of big data technology causes the data from agriculture domain to enter into the big data. They are not conventional techniques in existence to process such a large volume of data. The processing of large datasets involves parallel computation and analysis model. Hence, it is necessary to use big data analytics framework to process a large image datasets. In this paper, an automated big data framework is presented to classify the plant disease condition. This framework consists of a series operations that leads into a final step. When the classification is carried out using novel image classifier. The image classifier is designed using a Convolutional Recurrent Neural Network Classifier (CRNN) algorithm. The classifier is designed in such a way that it provides classification between a normal leaf and an abnormal leaf. The classification of plant images over large datasets that includes banana plant, pepper, potato, and tomato plant. Which is compared with other existing big data plant classification techniques like convolutional neural network, recurrent neural network, and deep neural network, artificial neural network with forward and backward propagation. The result shows that the proposed method obtains improved detection and classification of diseased plants compared to other the convolutional neural network (94.14%), recurrent neural network (94.07%), deep neural network (94%), artificial neural network with forward (93.96%), and backward propagation method (93.66%).

Featured Image

Photo by Alex Block on Unsplash

Photo by Alex Block on Unsplash

Why is it important?

In this paper, the authors present an automated big data framework to classify the plant disease condition. This framework consists of a series operations that leads into a final step of CRNN approach, where the classification is carried out using novel image classifier. The image classifier is designed using a CRNN Classifier algorithm. The classifier is designed in such a way that it provides classification between a normal leaf and an abnormal leaf. The analysis is carried out on healthy and unhealthy leaves of pepper, banana, tomato, and potato. The classification of plant images over large datasets is compared with other existing big data plant classification techniques and normal classification techniques. The result shows that the proposed method obtains improved detection and classification of diseased plants. In future a study of maybe uses optimized histogram-based and U-net based feature extraction is suitable for both binary and multi-class leaf diseases classifications.

Perspectives