Downloads: 7 | Views: 394 | Weekly Hits: ⮙1 | Monthly Hits: ⮙1
Experimental Result Paper | Computers & Electronics in Agriculture | Ethiopia | Volume 12 Issue 9, September 2023 | Popularity: 5.4 / 10
Avocado Leaf Disease Classification Using Convolutional Neural Network
Hirut Abera, Varagantham Anitha Avula
Abstract: Countries agricultural production of plants and crops were affected by pests and pathogens. Avocado is one of the popular subtropics fruit of high economic importance. Avocado plant has so much importance for food consumptions, oil production and financial source of income, however most of the time it is affected by the different diseases that result in the reduction of quality and quantity of the avocado production. Usually avocado disease symptoms were appear on the leaf of the plant and through time it affect the fruit and stem of the tree leading to the death of the plant. Traditionally farmers and specialists or experts detect and identify plant disease by naked eyes. This method was inaccurate and expensive. Detection using image processing technique will be accurate and fast. In this thesis the researcher proposes deep learning based avocado leaf disease detection and classification. Algal leaf spot, leaf burn, perseamite damaged and powdery mildew disease of avocado that appear on leaf of avocado plant were concerned. This research has passed through the following steps, image acquisition, image preprocessing, feature extraction, training the model and detection and classification. The proposed method was classified into three phases, by changing the values of hyper parameters. All phases had passed the above processing steps. Avocado leaves affected by four diseases and healthy leaves were trained and classified into their disease class. To test the model performance, data set was split into training set and validation set in the ratio of 80%, 20% respectively. Convolutional Neural Network (CNN) model is used for the detection and classification purpose. CNN has shown good performance in training, prediction of the disease and classification that result 100% for training accuracy and 98.8% for testing accuracy with the overall accuracy of 99%.
Keywords: Avocado Disease, Convolutional Neural Network, Deep Learning
Edition: Volume 12 Issue 9, September 2023
Pages: 780 - 785
DOI: https://www.doi.org/10.21275/SR23908191606
Please Disable the Pop-Up Blocker of Web Browser
Verification Code will appear in 2 Seconds ... Wait