The conventional methods adopted by the formers for leaf disease detection and classification can be monotonous and unreliable. It is challenging for formers sometimes to attempt and anticipate the type of disease manually. The inability to early diagnose the disease and erroneous predictions may da...
Kaydedildi:
| Asıl Yazarlar: | , , |
|---|---|
| Materyal Türü: | Makale |
| Online Erişim: | https://doaj.org/article/3a3cbeef0f614dc3ba27bb9e68cc4892 |
| Etiketler: |
Etiketle
Etiket eklenmemiş, İlk siz ekleyin!
|
| Özet: | The conventional methods adopted by the formers for leaf disease detection and classification can be monotonous and unreliable. It is challenging for formers sometimes to attempt and anticipate the type of disease manually. The inability to early diagnose the disease and erroneous predictions may damage the crop, resulting in loss of crop production. To prevent losses and increase crop production, computer-based image classification methods can be adopted by the formers. Several methods have been suggested and utilized to predict crop plant diseases using pictures of unhealthy leaves. Investigators are currently making significant advances in the detection of plant diseases by experimenting with various methodologies and models. Artificial Neural Networks (ANNs) stand out as a widely employed machine learning method for effectively classifying images and predicting diseases. Alongside ANNs, other prevalent algorithms include Linear Regression (LNR), Random Forest Algorithm (RFA), Support Vector Machine (SVM), Convolutional Neural Networks (CNN), and k-nearest Neighbor (KNN). Combining these algorithms has been explored in various studies to enhance accuracy. This review examines their application in classifying diseases in citrus crop leaves, focusing on metrics like Accuracy, Precision, and Sensitivity. Each algorithm has its strengths and weaknesses in disease identification from leaf images. The accuracy and effectiveness of these algorithms depend significantly on the quality and dimensionality of the leaf images. Therefore, a reliable leaf image database is crucial for developing a robust machine-learning model for disease detection and analysis. |
|---|