Main Article Content
Object Detection (OD) in surveillance video is the way of automatically detecting and tracking object classes of interest within the video recording. It includes the application of a Computer Vision (CV) technique to analyze the video frame and identify the classes of objects or the presence of specific objects. Various OD techniques are used to find objects within the footage video. This algorithm analyzes the visual feature of the frames and employs Machine Learning (ML) approaches namely Deep Neural Network (DNN), to detect and track objects. It is worth mentioning that the accuracy and performance of OD in surveillance video depends on factors including the choice of algorithms and models, the availability of labelled training data, and the quality of the video frame for the specific object of interest. This study introduces a new modeling of Intelligent Object Recognition and Classification by employing Aquila Optimizer with Deep Learning (IODC-AODL) approach in Surveillance Video. The goal of the IODC-AODL technique is to integrate the DL model with the hyperparameter tuning process for object detection and classification. In the proposed IODC-AODL approach, a Faster RCNN method is enforced for the process of OD. Next, Long Short-Term Memory (LSTM) networking approach is implemented for the object classification process. At last, the AO approach is enforced for the optimum hyperparameter tuning of the LSTM network and it assists in improving the classifier rate. A widespread simulation sets are performed to exhibit the superior performance of the IODC-AODL approach. The experimental result analysis portrayed the supremacy of the IODC-AODL algorithm over other models.