Vision based flying vehicle tracking

In this paper, we have proposed an efficient vision based method for the tracking of flying vehicles via a moving video camera. The suggested approach is based on the combination of the feature extraction and matching algorithm as well as the 2D deformable mesh surfaces. We have developed a new set of energy functions for mesh to improve the performance and quality of the tracking algorithm. The proposed method is capable of estimating both the kernel and contour of the flying vehicle. To evaluate the suggested system, we have defined a criterion to measure the tracking error percentage (TEP) during video frames. The proposed method tested with several flying vehicles videos in different atmospheric conditions and maneuvering motions and experimental results showed high accuracy, stability, and convergence in noisy and noiseless video sequences.