Get Cure: Machine Learning-Based Plant Disease Detection

Abstract

Early diagnosis of plant diseases that impact agricultural productivity rate is not possible in the conventional farming business due to a lack of technological tools. Earlier work applied Machine Learning (ML) algorithms were employed to improve accuracy in detecting plant disease. However, the recent development of Deep Learning (DL) was encouraging with regard to accuracy and dependability. Our study intends to enhance the precision as well as the dependability of plant disease symptom diagnosis using a Convolutional Neural Network (CNN) and a Deep Neural Network (DNN). Additionally, it provides remedial actions and supplements for the affected crop. We implemented the Convolutional Neural Network because of its ability to capture local patterns and spatial relationships in the input data, making them potential for image analysis tasks. The Deep Neural Network can extract high-level features due to the performance of non-linear transformations and also the detection process was more efficient than traditional one. The transfer learning algorithms have improved the accuracy of our model in plant disease detection and recognition, thereby aid to improve the productivity of crops. In our proposed work, CNN architecture along with alternative models including VGG and ResNet, is trained on the dataset and the proposed ResNet model improves the accuracy.

Abstract from the paper’s OpenAlex record, shown for indexing and research. Copyright remains with the authors or the publisher. Read at the source ↗

The text the model read, with the source record’s inline markup (italics, sub- and superscripts) rendered and its character entities decoded; quote offsets refer to the raw text.