Visual MAV Tracker with Adaptive Search Region

Tracking a micro aerial vehicle (MAV) is challenging because of its small size and swift motion. A new model was developed by combining compact adaptive search region (SR). The can accurately robustly track MAVs with fast computation speed. SR, which slightly larger than target MAV, less likely to include distracting background large SR; thus, it the MAV. Moreover, SR reduces time tracking be conducted relatively shallow network. An optimal MAV ratio obtained in this study. However, causes frequent failures presence dynamic proposed address problem; adaptively changes location based on size, location, velocity SR. without strategy tracks an accuracy 0.613 robustness 0.086, whereas has 0.811 1.0. online accomplished within approximately 400 frames per second, significantly faster real-time