Optimal Target-Oriented Fusion of Passive Millimeter Wave Images with Visible Images Based on Contourlet Transform

Passive millimeter wave (PMMW) imaging system forms images through passive detection of millimeter wave radiations emitted from the objects. This system allows detection of invisible characteristics of a scene. The important characteristic of these waves is their penetration into cloth fibers and bad climatic conditions such as mist, rain and smoke. These waves are able to detect objects (particularly metal objects) hidden in cloths. For this reason, imaging system is applied in security section of the airports. This system has also abundant applications in military, meteorology, navigation industries in low vision and medical industries. In airports and generally the places where there is need for physical inspection of people, millimeter wave scanners are used. These devices are able to detect weapon and other objects hanging on body of the person or hidden in clothes. Due to limitation of dimensions of lens and antenna and limitation of light refraction, images obtained from this system have low resolution. Imaging in this system acts as a low-pass filter. For this reason, high-frequency information of these images is removed; therefore, set of actions is necessary for improving quality of these images. Different methods have been presented for increasing quality of these images which are divided into two classes of retrieval and combination or fusion of image (for example, fusion of PMMW image with visible image). Target of this paper is to present a new method for fusion of millimeter wave images with their visible images with their visible images with contourlet transform and a new method in thresholding; so that the output image includes visual information very close to visible image which contains object hidden from millimeter wave image.