A Hierarchical Model Based UAV Target Tracking Method

Abstract

The invention discloses a UAV target tracking method based on a layered model, which belongs to the field of UAV remote sensing image processing, and comprises the following steps: the first step is to establish the feature expression of the initial frame target; the second step is to establish the mean value Drift-based coarse tracking model; the third step is to establish a fine tracking model based on improved particle swarm optimization. By introducing a layered model combining mean shift and improved particle swarm optimization, this algorithm effectively overcomes the problem of difficult, inaccurate and easy-to-lose video tracking due to the particularity of UAV imaging, and effectively improves the accuracy of target tracking.

Description

technical field [0001] The invention belongs to the technical field of remote sensing image processing, and in particular relates to a method for tracking an unmanned aerial vehicle target based on a layered model. Background technique [0002] Unmanned aerial vehicles (UAVs) are used more and more widely in both civilian and military fields, so their development has attracted the attention of various countries. Among many research directions, visual capabilities will become the focus of future development and competition of UAVs. UAV visual tracking, as the main force of UAV vision, has also become a key technology. According to public information, as early as 2005, the computer vision laboratory of the University of Central Florida in the United States developed the COCOA system based on MATLAB, which is mainly used for target detection and tracking processing of low-altitude aerial video images of UAVs. At present, domestic target tracking methods can be roughly divided ...

AI Technical Summary

Problems solved by technology

The advantage is that the search speed is fast, the search efficiency is high, and the algorithm is simple; the disadvantage is that it does not handle discrete optimization problems well, and it is easy to fall into local optimum

[0006] Among them, the particle filter strategy and the particle swarm optimization search strategy both need a certain number of particles. When the number of particles is more, the accuracy rate will be higher, but the search time will be correspondingly increased; when the number of particles is small, the search time will be reduced. Small, but some areas may not be searchable, because the particle swarm optimization strategy uses the iterative idea, the impact on the accuracy is small, but the particle filter will seriously affect the tracking accuracy

[0007] Due to the particularity of the imaging environment of the UAV, a relatively stable video image can be obtained during the smooth flight process. However, due to the interference of the gimbal movement and airflow, the UAV may be subject to certain impacts during the flight. In this case, the target may have a large motion offset between the resulting image frames

[0008] From the above analysis, it can be seen that the existing target tracking method based on the UAV reconnaissance platform has a contradiction between accuracy and efficiency, and it is difficult to meet the needs of real-time and high-precision target tracking in the modern battlefield.

Images