Development of a Surveillance System for Red Tilapia Anomaly with Technology Neural Network
Abstract
The objective of this research is to develop a surveillance system for red anomalies. Tilapia in the cage with neural network technology by taking photos of Red Tilapia in the cage, Mueang District, Phitsanulok Province which is classified into 2 types, normal and dead, 10,800 images.
Still Image and Video Processing Red Tilapia number 10 videos. Video tracing using Faster R-CNN model for image classification. The accuracy in the still image test is 95.37 %. And the video tracking accuracy is 94.79% and the notification system is created through the LINE application