Passive Millimeter Wave Imaging Super Resolution Algorithm Research And Implementation

The imaging system using of millimeter wave can penetrate through the clothes and with no radiation, So it will not be a threat to the security of the detector, while the detection will not feel the privacy is violated. In addition, With the resolution of millimeter wave radiation brightness temperature meter increases, it is easy to separate dangers metal, some non metal targets and human body, so the passive millimeter wave imaging system in screening alarm, fixed security and other fields has been applied more and more widely. But because of present manufacturing process etc, the channel array can not be placed too densely, which leads to the low resolution of imaging results and can not meet the practical application. Therefore, it is needed to study the method to improve the spatial resolution of the millimeter wave image so as to satisfy the needs of the users.The research content of this paper is based on the practical scientific research project. Firstly, the imaging system and the reduced model of the millimeter wave image was studied, and then the system and the calibration scheme was analyzed. After the imaging results are presented, a millimeter wave super resolution algorithm based on statistical optimization and integral variation is proposed, and the millimeter wave super resolution algorithm based on partial differential equations is also studied. Specific research contents include1. The theory of passive millimeter wave imaging and detection theory were studied, and the image degradation model in imaging process was analyzed.2. The RL(Richardson-Lucy) algorithm based on statistical optimization and the super resolution restoration algorithm based on TV(total variation) are studied, then proposed a algorithm combined the two super resolution restoration algorithm for noisy millimeter wave image and achieved good results.3. The super-resolution algorithm based on partial differential equations and total variation, including the standard total variation restoration(ROF) model and improved total variation restoration model were analyzed, based on this,we analyzed the superresolution restoration based on high-oder and pTV-L, then a combination of high-order model and the ROF model super resolution restoration algorithm is proposed.4. Study on TV-Stokes model super resolution restoration algorithm,for the poor results of sharp edges image, an improved model based on the basic TV-Stokes model is studied. The energy functional is reconstructed in the second step of the TV-Stokes model, and the defects of the basic TV-Stokes model are eliminated.5. One of a passive millimeter wave imaging system of detecting engineering implementation and experiment are summarized and discussed。...