Research On Background Suppression Method Of Infrared Small Target Under Cloud Background

With the development of image processing technology, infrared target detection technology is widely used in many areas. Implement background suppression of infrared weak small targets that under clouds background gradually become a research hot spot at home and abroad. But with the characteristics of infrared weak small targets, combined with cloud and noise interference, target is easy to integration in the cloud background image. It brings great difficult and challenge for the background suppression of infrared weak small targets under clouds background.Based on this,the thesis mainly do several works as follow:Firstly, this article research the characteristics of infrared weak small targets image,Mainly analyze the characteristics of clouds background, the sources of noise and the characteristics of infrared weak small targets. By analysis the characteristics of the various components that in the infrared images. Pointing out the factor that influenced background suppression of infrared weak small target. It will provide solid theoretical support for background suppression of clouds background infrared image.Secondly, this article deals with three kinds of typical background suppression method at present, They are median filtering, morphology filtering and wavelet transform,According to one of the more active pulse median filtering, multiple structure morphology filter and wavelet threshold transform for analyzed the two infrared images that the target position in different clouds.Thirdly, in view the disadvantages of the traditional wavelet threshold image background suppression method, we are improved its threshold function and wavelet, the author bring an improved wavelet threshold algorithm, this algorithm not only could keep target information as much as possible,But also can effectively suppress noise effectively.The results show that the improved algorithm can better improve the noise rat.Comparing with three background suppression methods, the MATLAB simulation experiment has confirmed that the improve wavelet threshold transform methods in this article have advantages in the feasibility and the superiority.