Abstract:The air‐to‐ground visual object tracking has several applications, including surveillance, cinematography, and chasing. Briefly, camera states and a vision‐based range estimation are added to the tracking method to locate the target in inertial coordinates and introduce a probability distribution to predict the future positions of the target. The results of adding this motion model to the DiMP tracker demonstrate a 19.2% tracking precision improvement. Visual object tracking, crucial in aerial applications such as surveillance, cinematography, and chasing, faces challenges despite AI advancements. Current solutions lack full reliability, leading to common tracking failures in the presence of fast motions or long‐term occlusions of the subject. To tackle this issue, a 3D motion model is proposed that employs camera/vehicle states to locate a subject in the inertial coordinates. Next, a probability distribution is generated over future trajectories and they are sampled using a Monte Carlo technique to provide search regions that are fed into an online appearance learning process. This 3D motion model incorporates machine‐learning approaches for direct range estimation from monocular images. The model adapts computationally by adjusting search areas based on tracking confidence. It is integrated into DiMP, an online and deep learning‐based appearance model. The resulting tracker is evaluated on the VIOT dataset with sequences of both images and camera states, achieving a 68.9% tracking precision compared to DiMP's 49.7%. This approach demonstrates increased tracking duration, improved recovery after occlusions, and faster motions. Additionally, this strategy outperforms random searches by about 3.0%.

Comparison of Dynamic Models for Aerial Target Tracking Maneuvers Based on Stability and Measurement Loss

Classify Motion Model via SVM to Track Underwater Maneuvering Target

Survey of maneuvering target tracking. Part I. Dynamic models

Data-Driven Approaches for Modelling Target Behaviour

Applying Dynamic Model for Multiple Manoeuvring Target Tracking Using Particle Filtering

Intelligent Tracking Method for Aerial Maneuvering Target Based on Unscented Kalman Filter

Unified Maneuvering Model For Target Tracking By Using Range-Rate Measurement

Deep Uncertainty-aware Tracking for Maneuvering Targets

Adaptive Modified Unbiased Minimum-Variance Estimation for Highly Maneuvering Target Tracking With Model Mismatch

An Adaptive Single Model of Maneuvering Target Tracking

Reference Acceleration Based Dynamic Model for Maneuvering Target Tracking

Dynamic Based Estimator for UAVs with Real-time Identification Using DNN and the Modified Relay Feedback Test

Probabilistic 3D motion model for object tracking in aerial applications

Adaptive Distributed Multiple-Model Filter with Uncertainty of Process Model

Reference acceleration-based dynamic model for maneuvering target tracking

Adaptive Interacting Multiple Model Algorithm Based on Information-Weighted Consensus for Maneuvering Target Tracking.

Passive Tracking of Maneuvering Targets Based on Current Statistical Model

A Survey of Manoeuvring Target Tracking Methods

A Novel Approach to Maneuvering Target Tracking Based on Random Motion Model Using Random Kalman Filtering

Application of Adaptive Weighted Strong Tracking Unscented Kalman Filter in Non-Cooperative Maneuvering Target Tracking

The Design of State Estimator of Highly Maneuvering Targets from Measure-only Line-of-sight Angular Velocity