Research On Millimeter Wave Dim Small Target Detection Method In Complex Background

Millimeter waves dim small target detection is a relatively new research field.The detection of dim small targets using millimeter wave has unique advantages compared with radar,visible light,infrared and other technical means,which has very high application value in the military field such as remote investigation and air defense early warning.It is a challenging task to detect the dim small targets in the complex background with long distance,there is no good solution yet.Therefore,it is very important to combine the characteristics of millimeter wave to detect small targets with high accuracy in order to improve the detection accuracy of small targets.In this thesis,the dim small target images of millimeter wave are studied,and three dim small target detection algorithms are proposed.According to the characteristics of passive millimeter wave imaging and the difference between small target and background clutter,a kind of dual dictionary of target and background detection algorithm is proposed.Firstly,an over-complete dictionary is constructed according to the content of the image itself,and then the original sparse representation is improved in order to complete classification of target and background over-complete dictionary.Finally,the small target is detected by sparse coefficient difference between the target and the background.Background suppression method is a common method in small target detection.In this paper,based on the human visual system(HVS)and the combination of the target region and the local background characteristic prior,we propose a background suppression algorithm based on circle-surround center difference.In this algorithm,a circle center difference model is designed,which can achieve the purpose of background suppression by cumulative difference of the pixel values between the central region and circular region.The algorithm not only can effectively distinguish the target area and the background area,but also can eliminate the influence to the directional clutter region very well.For passive millimeter wave images' problems of small target and background region difference is weak,and the feature of target region can be extracted is small,combined with background target accurate dual dictionary detection algorithm and circle center difference background suppression algorithm,a new kind of algorithm by improving the sparse representation and circle-surround center difference model is proposed.The algorithm combines the advantages of two kinds of algorithms for joint background suppression,which results in better enhancement effect for target and better background suppression.The experimental results show that the detection accuracy of the proposed algorithm is higher than that of other mainstream algorithms under the same conditions.