Font Size: a A A

Passive Millimeter Wave Imaging Super Resolution Algorithms

Posted on:2013-07-30Degree:MasterType:Thesis
Country:ChinaCandidate:Y F ChenFull Text:PDF
GTID:2248330374986526Subject:Signal and information processing
Abstract/Summary:PDF Full Text Request
Passive millimeter wave imaging (PMMWI) makes use of the differences in the distribution of millimeter-wave radiation energy of different targets in the scenarios to achieve imaging. The passive millimeter wave imaging has the unique ability of high penetrability, strong counter camouflage, good concealed performance and so on, which can do real-time monitoring at all weather and all time. Therefore, it is widely applied to airport security, scene monitoring, traffic navigation, etc. but for the system low pass effect, the millimeter wave image acquired has low resolution.Therefore, it is particularly important to research super-resolution reconstruction to improve the resolution of acquired millimeter wave image. Supported by National Natural Science Foundation of China, super-resolution reconstruction for passive millimeter wave imaging has been explored and investigated in this dissertation, and the main contents are as follows:1. Using Richardson-Lucy(R-L) algorithm and accelerated R-L algorithm to enhance the millimeter wave image resolution, we use an improved way to enhance the accelerated R-L algorithm performance. The improvement is to use low frequency component restored by least square method to replace the low frequency component restored by original algorithm.2. In view of the discontinuity and noise amplification of the interpolated image in projections onto convex sets (POCS), we adopt Gaussian function to smooth the interpolated image, avoiding the enlargement of the discontinuity and noise during the POCS iterative process.3. In the POCS iterative process, the point spread function (PSF) can not effectively project back and compensate all points of low-resolution image. So, we improve the PSF so that all points can be back projected to the high-resolution image and enhance the resolution.4. Due to the fringe fold-over, the reconstructed result is not good. So, we process the edge detection before iterative correction, and according to the fringe and non-fringe, we use different PSF to project back all points in order to reduce the fold-over effectThe improved methods mentioned above have been verified with through analogous and real millimeter wave image through simulation experiments. The experiment results have proved the effectiveness and applicability of the reformative algorithms both subjectively and objectively.
Keywords/Search Tags:PMMWI, super-resolution, R-L, POCS
PDF Full Text Request
Related items