There is a recognized need to develop improved systems for survey and obstacle detection offering characteristics such as day, night and adverse weather capabilities, acceptable operating range and adequate spatial resolution. Passive infrared (IR) systems can perform well at night. They are also able to penetrate smoke and haze to a far greater distance than visible systems, but have difficulties when trying to detect through clouds, fog and light rain, since these attenuate strongly infrared radiation. In contrast, millimeter-wave imaging systems suffer little attenuation in either cloudy or foggy conditions, especially around the 35 GHz, 94 GHz and 140 GHz frequency band. This paper presents the development of a low cost one-channel radiometric system working around 52 GHz. The authors have chosen to use a noise-added radiometer with low energy consumption. Four important points are discussed in this paper. First, the radiometric visibility gives the investigation volume for different weather conditions. The computation of such volumes shows a real interest for the radiometric technique. Secondly, a new calibration method, based on a modulated noise diode source, followed by a real-time Kalman filter processing is validated on a 4 GHz radiometer. Thirdly the millimeter-wave radiometer in the 51 GHz - 55 GHz band is described. The authors present also the low-noise amplifier (LNA) that will be used in the final project of the imaging system working in the 94 GHz band. Finally the image processing, based on a deconvolution process and followed by a regularization of the temperature improves the spatial resolution and optimizes the radiometer matrix located in the focal plane of a Fresnel lens.
Abstract