Evaluation of parameters in a neural network for detection of red ring pest in oil palm

In Latin America, Ecuador, Colombia, and Honduras have the highest oil palm production, diseases that quickly spread in plantations affect this production. That is why it important for these countries to control over plagues. Using pesticides produces environmental pollution respiratory problems workers. Therefore, necessary apply precision farming technologies can prevent from having use when tree already dead, reduce increase quality of oil. The aim project evaluate different neural networks detection This parameter starts with type device data acquisition, convolutional network, weight trim rate, false alarm rate. design method was developed using incremental strategies. Finally, drone Haar, network a rate 90% detect disease 98% accuracy. conclusion, robot has better accuracy uses an LBP training positive negative images are same. However, its low compared Haar training.