| Infrared imaging technology has good concealment and anti-interference,and can adapt to complex climate change,so it has a wide range of applications in military and civilian fields,such as infrared guidance,early warning,missile tracking,intelligent security,night navigation and so on.Infrared dim-small target tracking is one of the key technologies in this filed,which plays a very important role in military defense.Since the signal-to-noise ratio and contrast of the infrared image are relatively low,the clutter interference of the background is very serious,and the infrared dim-small target itself has the characteristics of long imaging distance,weak target,fast moving speed,lack of texture and shape information,etc.In the process,it is extremely easy for the target to be submerged in the background and cause the target to be lost,which makes the tracking of infrared dim-small target extremely challenging.Therefore,how to achieve effective and stable tracking of infrared dim-small target under complex back-ground conditions is of great research significance.Based on the characteristics of infrared dim-small target images,this paper analyzes the reasons for the deviation of the infrared dim-small target tracking process,and proposes corresponding methods to improve the tracking deviation phenomenon.The specific research contents are as follows:Aiming at the problems of tracking deviation caused by the long imaging distance of the infrared dim-small target image,the lack of texture and shape information of the target,the large background proportion,and the serious clutter interference,this paper proposes a context-aware infrared dim-small target tracking algorithm based on multifeature fusion.The algorithm is based on KCF algorithm.In the feature extraction stage,two features are selected to train the filter,and the training results are weighted and fused to predict the position of the target,then the context-aware framework is introduced into the algorithm to sample the background information,and the sampling results are used as negative samples to train the classifier,so as to reduce the impact of the background on the tracking results.The simulation results show that the algorithm can better improve the tracking deviation.Aiming at the problems of tracking deviation caused by weak target and target jitter in complex background situations,this paper proposes a context-aware infrared dimsmall target tracking algorithm based on singular value decomposition and side window filtering.Based on the introduction of the context-aware framework,the algorithm first uses singular value decomposition to preprocess the infrared dim-small target image to enhance the target,then uses the box filter with side window filtering technology to suppress the background edge in the base sample of each frame in the image sequence,so as to reduce the influence of the background edge on the tracking algorithm.The simulation results show that the algorithm reduces the influence of the complex background on the tracking process of infrared dim-small target,reduces the deviation of the algorithm,and can achieve effective and stable tracking. |
