Research On Super-resolution Reconstruction Of Passive Millimeter Wave Image Sequence

The needs of current security put forward a new request to the corresponding technology,on the one hand,it needs fast and real-time detection of personnel to carry the hidden dangerous goods;on the other hand,it requires security system security,no radiation,security personnel’s privacy will be not violated.The passive millimeter wave(PMMW)detection imaging technology have been widespread concern because it meets the above security requirements.However,the diffraction limitation of the PMMW detection imaging system is serious,resulting in the lack of high frequency detail information of the PMMW image and low spatial resolution.Therefore,it is necessary to do super-resolution the image in the signal processing for improving the image quality and the system resolution.The key to the image sequence super-resolution processing algorithm is how to accurately extract similar information between frames,so as to reconstruct a high-resolution image from a low-resolution image sequence.Based on the actual PMMW detection imaging project,this thesis studies the super-resolution reconstruction technology of PMMW image sequence.The main research contents and achievements include:1、Based on the basic theory of PMMW detection imaging technology,I analyzed the image degradation model in the process of system imaging,studied and established the mathematical model of super-resolution reconstruction of PMMW detection image sequence based on projection on convex set.2、The optimization and improvement of the super-resolution algorithm for image sequence based projection on convex set(POCS)were studied.In view of the incompleteness of the iterative termination condition in the classical POCS algorithm,I studied the adaptive POCS algorithm based on entropy constraints according to the scanning characteristic of the PMMW detection imaging system and the base of the characteristics of the PMMW image’s analysis.By setting the entropy difference threshold,the whole constraint on the image in the reconstruction process is realized.The improved algorithm can effectively restore the original PMMW image’s loss details,thereby improve the quality of the reconstructed image.3、In order to solve the problem of stochastic selection of confidence parameters inPOCS algorithm,the speed and precision of reconstruction are adjusted by setting the decreasing step size factor related to entropy difference.It can not only improve the reconstruction effect,but also guarantee the convergence speed and reliability of the algorithm.Simulation experiments based on optical image and PMMW image show that the adaptive POCS algorithm based on entropy constraint is better than classical POCS algorithm in subjective visual effects and objective quantitative analysis,and is suitable for PMMW imaging.4、In order to solve the problem of multi-frame POCS super-resolution algorithm,the computational complexity increases with the increase of resolution magnification.According to a recursive multi-layer processing idea,this thesis proposes a high-resolution image reconstruction based on POCS algorithm.All low-resolution images are reconstructed using the displacement relationship between frames,and POCS super-resolution image reconstruction is performed for each group.The results of theoretical and simulation experiments show that the fast algorithm can greatly improve the operation speed.