Arecanut Disease Detection
1Roshan S, 2Chethan B, 3Srushti M, 4Darshan V, 5VidhyaShree NR
1UG Student Dept . of CSD, 2UG Student Dept . of CSD, 3UG Student Dept . of CSD, 4UG Student Dept . of CSD, 5UG Student Dept . of CSD
1,2,3,4,5 Presidency University, Bengaluru
ABSTRACT :
The tropical crop arecanut, sometimes referred to as betel nut, is primarily farmed in India. In terms of arecanut production and consumption, the nation ranks second in the world. The areca nut plant is vulnerable to numerous diseases that impact its roots, stem, leaves, and fruits throughout its life cycle. While some of these illnesses can be seen with the naked eye, others cannot. These illnesses are brought on by abrupt changes in temperature and other meteorological factors; early disease identification is crucial. In order to minimize losses for farmers, this work focuses on early and precise disease diagnosis. Using convolutional neural networks, we developed a system that assists in identifying arecanut, leaf, and trunk ailments and offers treatments. A Convolutional Neural Network (CNN) is a Deep Learning method that uses an image as input, gives different items in the image learnable weights and biases, and then uses the results to determine which objects are different. We took a dataset with 620 photos of arecanuts in both good and unhealthy conditions in order to train and evaluate the CNN model. An 80:20 ratio is used to separate the test and train data. Adam serves as the optimizer function, accuracy serves as a measure, and categorical cross-entropy serves as the loss function for the model compilation. To attain high validation and test accuracy with little loss, the model is trained over a total of 5 epochs. It was discovered that the suggested method was successful and 88.46 percent accurate in detecting arecanut illness. Diseases frequently seen in areca trees include Mahali Disease (Koleroga), Bud Rot Disorder, Stem Exudation, Yellow Leaf Blotch,Yellow Disease, which arises from persistent rainfall and climate alterations, these ailments need to be managed in the initial phase of infection; otherwise, it could lead to difficulties in oversight in the concluding phase that could result in detriment to the latter. To prevent this, we can utilize Machine Learning for disease detection. We will identify yellow leaf spot, stem bleeding, and Mahali disease (KoleRoga) in this project and provide treatments for the conditions we find.
Key words: Convolution Neural Networks, Arecanut, and Machine Learning
