A smart agricultural application: automated detection of diseases in vine leaves using hybrid deep learning

This paper reports a study which utilizes deep learning for automated detection of the symptoms diseases on vine leaves. Vine fruits or grapes are very important and have existed in Syria surrounding areas (e.g., Turkey) many years. Quality is also grape production as it consumed these every day. The aim this to improve diseasedetection accuracy leaves develop system help Syrian Turkish farmers agricultural engineers maintain quality production. In study, over 1000 images been collected from yards internet. These processed using MATLAB 2018B, Deep Learning Toolbox including convolutional neural networks (CNNs) with AlexNet, GoogleNet ResNet-18. A standard transfer (TL) algorithm used CNNs, whereas multiclass support vector machine (SVM) whilst GPU CUDA accelerating process disease software has created that enables automatic efficient nine types leaf identification healthy Experimental studies showed total reaches 92.5%, 87.4% 85.0%, 85.1% when AlexNet+TL, ResNet-18+TL, GoogleNet+TL AlexNet+SVM respectively. smart application can provide early thus fruits.