Durian Farm Threats Identification through Convolution Neural Networks and Multimedia Mobile Development

• Main Authors: Yusof, Aiman, Kamarudin, Noraziahtulhidayu, Al-Emad, Nabil Ali, Sapuan, Khusairi
• Format: Article
• Language: English
• Published: Ijetea 2023
• Subjects: T Technology (General)
• Online Access: http://eprints.uthm.edu.my/8899/1/J15817_93d696d741ce66312d4270d55ad734db.pdf; http://eprints.uthm.edu.my/8899/; https://doi.org/10.46338/ijetae0223_02

The difficulties to drive away the durian farm threatens animals such as wild boars, monkeys, foxes, and squirrels during nighttime often experienced by durian farmers. Therefore, the Pro Durian application is proposed that allows farmers to identify durian threats through a camera phone with an alert feature activation when the system detects an animal to drive away those animals. The application implements a deep learning algorithm of Convolutional Neural Network (CNN)-YOLO3in order to receive the best output results in identifying the different datasets of durian farm threats. The classification accuracies reached 80% in detecting the animal’s images.