Robust S-Y-biLSTM object tracking method for on-road objects shoot from an unmanned aerial vehicle

being one of the most challenging tasks in computer vision dynamical object tracking problems have faced additional issues regarding the unmanned aerial vehicle. In particular, image degradation, uneven object intensity, variety in object sizes, etc. In this paper, we proposed the implementation of the S-Y-biLSTM technique for on-road object shooting from an unmanned aerial vehicle. The proposed method contains the YOLOv4eff-based network as a feature map extraction, the SSDeff network as an object detection pipeline as well a bidirectional LSTM-based network as a tracking backbone technique. The object detection technique was improved by replacing the VGG-16 with the DenseNet-S-32-1. In addition, by suggesting a feature fusion module, we have added a residual prediction model for each feature layer within the SSD network for object detection. The training and testing process of the proposed network was performed based on the dataset of on-road objects shouted at the height of 15-45 meters. The results were shown that the proposed object tracking method S-Y-biLSTM has achieved higher accuracy and robustness compared to LYOLOv4eff, DeepSort and ROLO applied to the unmanned aerial vehicle.