Classifying Infected Palms with Dubas's Bug Based on Artificial Intelligence

In the agricultural field, many countries consider palm date a wealth and an important source of economy. Dubas bug is one of the insects that kill crops, reduce product quality, and cause yield losses. To detect and classify the infected palms, the conventional method is achieved by human vision. To automate the treatment of the infected palms, automatic Detection for the Dubas is required. This study presents an intelligent approach to detecting and classifying the infected palms by Dubas. The input data was collected by a drone camera and preprocessed by reducing the size of the data, then sliced then each image was sliced into several parts. Each part was decomposed using a lifting wavelet transform to reduce the data size. The preprocessed images are entered and trained by the convolutional neural network. The experimental results show that the accuracy of classification is above 97%.