DESIGN AND CONTROL OF ROBOTIC MANIPULATORS FOR PRECISION TASKS
Journal of Computational Analysis and Applications · 2025
The escalating human-wildlife conflicts necessitate the development of advanced monitoring systems to detect wild animal movements and mitigate potential threats.This paper proposes a hybrid deep neural network (DNN) model that integrates Convolutional Neural Networks (CNNs) and Long Short-Term Memory (LSTM) networks to accurately detect and classify wild animal movements.The system employs a CNN for feature extraction from video frames, followed by an LSTM to capture temporal dependencies, enabling real-time detection and classification.Upon detection, an alarm system is triggered to alert nearby personnel, enhancing response times and safety measures.The model is trained on a diverse dataset comprising various animal species, achieving an average accuracy of 98.5%.This hybrid approach demonstrates significant improvements over traditional methods, offering a robust solution for wildlife monitoring and conflict mitigation.