Utilizing Extreme Gradient Boosting (XGBoost) for Apple Scab Disease Detection

The research in this paper aims at evaluating a technique known as Apple Scab Disease detection using XGBoost (ASD-XG) Extreme Gradient Boosting (XGBoost) in the identification of apple scab on the apples crop which is a detrimental disease globally. The work feeds on a collection of 100 high resolution images of apple leaves on which image processing methodology is used to derive features pertinent to scab condition including color, texture and shape. The above-mentioned features of the dataset have been used to train the XGBoost model on the dataset, which showed a test accuracy of 95% as against the models like SVM, RF, CNN etc. Consequently, the experimental evaluations prove XGBoost’s higher accuracy and readability for early discovery of diseases in agriculturally important crops. The work done in this study indicates that XGBoost could underline the improvement of sustainable farming.