Line Pattern Removal and Enhancement Technique for Multichannel Passive Millimeter Wave Sensor Images

Passive millimeter wave (PMMW) imaging systems have attracted an increasing interest over the past years due to their superior poor-weather performance compared with visible and IR systems. In passive imaging, the spatial information acquired is strictly band-limited. A major drawback to PMMW images is their poor angular resolution. Also, another problem of singlechannel PMMW imaging systems is the slow response time due to the lack of thermal sensitivity. The imager could operate at TV (television) rates using a number of parallel channels which may reduce the extent of this problem. In multi-channel, differences between the responses of individual channels can introduce noise into the image, which can obscure details of interest. The proposed noise-removal technique is a two-pass combined method of two techniques. One technique is for removing the DC component in the frequency domain and the other one is statistical filtering based on homogeneous region in the image (spatial) domain, followed by high-boost filtering by 3×3 mask for enhancement the image. High quality images are presented to demonstrate the potential of this technique.