Abstract

Passive millimeter-wave (mmW) imaging has many applications specific to defense, security and safety since mmW wavelengths penetrate common obscurants such as haze, fog, smoke, dust, sandstorms, clouds, fabrics, or other thin dielectrics. For terrestrial remote-imaging applications, there are several primary bands of interest centered at 35, 94 and 135 GHz, defined by atmospheric transmission windows. A mobile mmW imaging system was constructed for use at 35 GHz utilizing an optical upconversion technique prior to my tenure as a graduate student, but the system suffered from performance issues. This system was less automated than desired, and imaged in only the horizontal polarization. Imaging with radiometric temperature data from both polarizations would allow for better identification of the scene being imaged, as well as permit the recognition of subtle features which were not previously apparent.

Additional design changes and enhancement of this system was necessary before completing the construction of the next generation of mmW imager for use at 94 GHz. The goal of the work of my research was to digitize signal acquisition and processing, improve spatial resolution and registration of the scene being imaged, increase portability of the system, reduce scan times, and enable additional capabilities. To this end, this thesis describes the designs which I implemented for imaging at these mmW frequencies which include: the enhancements made by digitizing the system; advances in vision, targeting, motion control and signal processing algorithms; automation of data collection and storage; and the addition of the abilities to perform radiometric and cross polarization imaging. The mmW imagers utilize a single-pixel rasterizing technique, and in order for these systems to perform well, the computational time to render each pixel needs to be small and consistent. The most significant contribution that I made was the development of a novel digital lock-in algorithm utilizing a closed form least-squares method. This refinement reduced the pixel computational overhead, thereby decreasing scan time, while resolving noise and spatial registration problems inherent in the previous implementation.

The goals of my research were met through the application of hardware changes and the coding of new software. The systems now utilize digital signal acquisition and processing, exhibit improvements in imaging metrics while benefiting from a reduction in total scan time, are more portable and robust in construction, and data is now automatically gathered and stored. The addition of temperature calibration and cross polarization hardware and software allows for the concomitant collection of spatially resolved radiometric scene images in both polarizations. These imaging systems were created to provide proof-of-concept for the optical upconversion technique and to prototype the eventual development of a two-dimensional distributed array imaging system. Little is well-known about mmW phenomenology; therefore, these systems were also developed to explore the world at 35 and 94 GHz. The modular design I created on the newly constructed platforms serves to allow portability between imagers, and provides the foundation for the development of additional systems and capabilities in this area of research.