Research On Infrared Small Target Detection And Tracking In Complex Interference Scene

Infrared small target detection and tracking is widely used in video surveillance,remote sensing technology,precision-guided munitions,military defense and other fields.However,there are a variety of scenarios with complex backgrounds in practical applications,which usually leads to false alarms in the detection results.Besides,in the military field,the infrared guidance system not only need to track the target in complex scenes stably but also to have the ability to resist artificial decoy.Therefore,it is of great significance to resist the natural interferences and artificial interferences in complex scenes and to complete the detection and tracking of infrared small targets.This dissertation analyzes the target and interferences characteristics and studies the infrared small target detection and tracking problem in complex interference scenarios.The main contributions of this dissertation are as follows:For the problem of infrared small target detection with complex backgrounds,the features of the weak small targets,edge,corner and other background clutter in the infrared image are first analyzed.Then,the differences between target and clutter in gray scale distribution and structure size are summarized.Finally,an algorithm based on the most similar neighbor patch background estimation is proposed to complete infrared small target detection in complex scenes.In this method,the traditional pixel-based background estimation algorithm is promoted to patch-based background estimation algorithm,which can preserve the background clutter in the complex image,reduce the edge clutter in the target image,and reduce the false alarm rate in the detection results.The experimental results show that the proposed algorithm is capable of distinguishing infrared small target from edge in the process of the background estimation,and has higher detection rate and lower false alarm rate.For the problem of the false alarms and similar interferences in the process of infrared small target tracking,this paper analyzes the features and motion behaviors of the target and proposes a target pre-association algorithm based on adaptive weighting multi-features fusion filtering.By establishing the target features model and target trajectory chain,the algorithm could resist the influence of interferences that have feature difference from the target effectively.For the problem of infrared decoy interference during target tracking,the target and decoy behaviors from decoy appearing to disappear are firstly analyzed.Then,the characteristics of the infrared decoy interference behaviors are summarized.On the basis of the above analysis,an infrared decoy interference event detection algorithm using feature compression and minimum risk criterion is proposed.Finally,based on the above two algorithms,an anti-decoy infrared small target tracking algorithm based on event detection is proposed.The experimental results show that the proposed algorithm can not only detect the infrared decoy event accurately,but also select different tracking strategies to track the target when the interferences occurs.It can effectively resist multiple interferences in complex scenes and complete the real-time target tracking.