Millimeter-wave Passive Detection Imaging Super-resolution Algorithm And Dsp Implementation

Passive millimeter wave (PMMW) imaging system provides special advantages over the optical and Infra-Red (IR) systems, especially under the poor weather with frog, clouds and mist which are effectively transparent at millimeter wavelength. Due to the reflection of the sky radiation by the metallic, building, road surface and the airstrip, the PMMW images are often acquired with good contrast. Meanwhile, the millimeter wave can partly see through the nonmetallic materials, so it is easy to be hided. Then compared to the active imaging, the passive is more likely to survive on the battlefield. Besides, the passive imaging system is easy to escape from the radar's tracking. Consequently, PMMW imaging system is more and more popularly used and has attracted more concerns. The PMMW imaging technique will be a new research trend in the future.However, the acquired image is a low-pass filtered version with noise of the original scene because of limited aperture dimension, which leads to the problem of poor resolution. Though the resolution of PMMW images can be enhanced through increasing the aperture dimension, it will make a high cost and suffer some tech-limitations in the fact. Therefore, robust and efficient resolution enhancement algorithms which are called super-resolution algorithms are required to improve the resolution of PMMW images. Because of the complexity of algorithm, general DSP can not achieve large storage capacity, high-speed real-time signal processing.This thesis was completed following a specific PMMW imaging project, focusing on super solution theory of PMMW imaging and signal processing unit of PMMW detecting system. which are described as follows:1. Research on the statistics-optimized super-resolution algorithm for millimeter-wave images, Design a software toolbox for statistics-optimized super-resolution algorithm. Research on Weighted space super-resolution algorithm.Weighted space super-resolution algorithm has a clear edge-blocked defects when PSF is at big size, this paper uses circle border technology to solve the deficiencies very well.2. Achieve statistical optimization-based image super-resolution algorithm and weighted space of the image super-resolution algorithm for real-time simulation and algorithm Analysis.3. In order to achieve super-resolution algorithm for real-time signal processing, this paper has developed signal processor of PMMW imaging. The signal processor using ADI provided TS201 EZ-KIT evaluation board, achieving a super-resolution algorithm library based on statistics-optimized. Through two TS201 processor parallel computing in the signal processor, finally reach the successful implementation of the super-resolution algorithm real-time signal processing.4. Achieve the united hardware testing and system debugging of the signal processor unit.