Visible/Passive Millimeter Wave Image Fusion Base On Multi-Wavelet Lifting Scheme

With the development of science and technology and the improvement of people's basic necessities of life,more and more people choose to travel to enjoy life.Therefore the security problem has become a focus.Traditional methods generally use weapons X-ray detector and metal detector.X-ray detector has a health risk to the human body,the metal detector for detecting non-metallic dangerous goods can't.To solve the problem of limitations of traditional methods,the passive millimeter wave imaging is the current hot direction of security technology research.Passive millimeter wave detector to detect target radiation temperature of the object itself,and has strong penetrating power and can detect hidden items.But passive millimeter wave detector for image resolution is not high.Visible light sensor is extremely strong light sensitivity,imaging with high resolution and details,but don't show hidden targets.Fusion of the visible light image and the passive millimeter wave image can complement each other,to provide more accurate and reliable information for the security.It can solve the problem of health risk of traditional security means,both security effectiveness and safety.Therefore,this paper studies the visible/passive millimeter wave image fusion algorithm.In the third chapter,in view of the security of passive millimeter wave image resolution is low and the characteristics of high resolution optical images,puts forward a way to deal with visual images and passive millimeter wave images,basing on multi-wavelet lifting scheme.First,we can get two images low and high frequency components after pre-processed of multi-wavelet and lifting wavelet transform.Then,the low frequency and high frequency components are fused by using different fusion rules to get a new image.lhird,do the post-processing for the fused image.The experimental simulation results show that this method can increase the accuracy of distinguish human body hidden items,realized the fast recognition of hidden items.In the fourth chapter,the paper puts forward a way to deal with visual image and passive millimeter wave image,basing on Non-subsampled Shear let Transform(NSST)and lifiting wavelet transform(LSWT)for the security check.First,low and high frequency components of two images were decomposed with NSST;Next,the low-frequency components are decomposed by the LSWT to get the other componets which are fused basing on regional variance and compressed,fused and reconstructed by CS.Then,doing the lifiting wavelet inverse transform.Third,the first high-frequency components were be deal with Laplacian.Finally,the inverse NSST is used to get the final fused image.The experiment indicates that the method's fusion effect is better,and it can realize the fast recognition of hidden items.In this paper,the research content has certain actual application value.It can increase the accuracy of distinguish human body hidden items,improve the safety,and security guarantees for the residents.