<![CDATA[Automatic fish detection in video is a challenging task in the field of computer vision, which can be addressed using deep learning methods. This study proposes the use of the Faster Region-based Convolutional Neural Network (Faster R-CNN) to detect two types of fish, namely Manfish and Lemonfish, in video data. The dataset was constructed by extracting frames from video and processing them using the Roboflow platform. The model was trained and tested using pre-split training and testing sets. The training process was conducted over 40 epochs using the Adam optimization algorithm to improve detection accuracy. Model evaluation was carried out using several metrics, including Precision, Recall, mean Average Precision (mAP), Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and R-squared (R²). The results show that the model achieved a precision of 94% and an accuracy of 87% for the Lemonfish class, and a precision of 95% and an accuracy of 89% for the Manfish class. These findings indicate that the model is capable of accurately detecting fish, delivering high detection performance, and effectively recognizing objects in video frames.]]>
Copyrights © 2024
