Object Detection in Satellite Images with Hierarchical Collective Hunter Algorithm Optimized Ensemble Classifier

Object detection in remote sensing images plays an essential role in computer vision task, which significantly identifies the position and class of the located objects with arbitrary orientation information. Recently, object detection plays an important role in various applications such as Surveillance and security in airports, organizations, and so on. However, target tracking is a very challenging task in natural images to attain appropriate detection results together with they obtain certain issues such as limited availability of high-resolute images, background complexity, and so on. To avoid these limitations, developed a Hierarchical collective Hunter algorithm with a Deep convolutional Neural network and a Bidirectional long short-term classifier (2HC-DBTM) model, which improved the interaction range of feature extraction by reducing the background complexity. Additional attributes mechanism such as the Hierarchical Collective Hunter (HCH) algorithm and hybrid attention mechanism are integrated with the proposed model for improve the ability and performance with minimum computational complexity and achieves better detection accuracy. The experimental results obtained in the evaluation of the model are achieved with high robustness and efficiency when compared with other conventional methods. The achieved performance evaluation of the model for evaluation metric F1-score is 97.71%, precision is 97.16%, recall is 98.27% and accuracy is 98.05%.