Research On Weak Aerial Target Tracking In Complex Cloud Background

Weak Aerial Target (WAT) detecting and tracking technology is related to the field of computer vision, image processing and pattern recognition. This paper has designed a automatically tracking system based on the WAT tracking, which was under the complex cloud background. Also verified its effectiveness through the experiment.First and foremost, this paper has summarized the significance, research status and hot issues in WAT tracking, which is under the complex cloud background. It has pointed out the importance in tracking by some typical application in home and abroad. Besides, this paper has analyzed the status of tracking method, and studied on the difficulty in WAT detecting and tracking. Last but not least, this paper proposed a tracking algorithm based on prediction automatically, and designed a general system for WAT. At last the algorithm has turned out to be effective through verification. In the aspect of WAT detecting, this paper has introduced the characteristics of Contourlet detection in multi-scale geometric analysis, discussed the multi-direction in Laplacian pyramid and multi-scale decomposition in Directional filter bank. Then extracted the energy features as basis of characterization from the sub-graph in WAT, and calibrated the WAT by principle of centroid, that has laid the foundation for tracking.In the aspect of WAT tracking, this paper has proposed a tracking algorithm based on prediction. According to the deficiency of tracking stability, this algorithm has combined with Contourlet detection to improve semi-automatic tracking in Mean Shift. Aiming at the movement of WAT in complex movement, this paper has introduced Kalman to establish tracking model when WAT was in the cloud cloak. It not only has better effect in tracking for low contrast WAT or attitude variation in forecast according to modify the model by centroid, but also solved the difficulty problem in tracking which can be resulted from occlusion tracking.This paper has verified the rationality of the system in WAT detecting and tracking through experiment simulation. In addition, this paper has pointed to the development of WAT tracking. This kind of analysis and algorithm has some reference meaning to WAT tracking system even to some complex system.