| abstract |
In one embodiment, an efficient method is presented for aerial searching for a small thermal target in a search area, such as a single person in open water, using two thermal imagers or “cameras” coupled with a computer which presents data from the system to a human user for inspection. One of the two thermal imagers has a very wide field of view (WFOV) fixed forward of or below the aircraft. The other, narrow field-of-view (NFOV) imager has a high zoom capability but its field of view can be reoriented to geo-point to a location on command. The WFOV thermal imager collects images rapidly so that no individual image is blurred due to changes in the field of view (FOV) on the time-scale of the image capture. The images are geo-registered using information from a global positioning receiver as well as the current altitude, roll, pitch, yaw, and velocity of the aircraft. As the aircraft moves and the FOV in the WFOV thermal imager changes, the computer averages the amplitude of the thermal radiation detected from each geo-registered position on the water below using the captured images continuously and in real time. The signal from a thermal target in the water is integrated while the background is relatively suppressed, enhancing the signal-to-noise ratio for the target as the square root of the number of images collected in which the target appears. A target which is much smaller than the area covered by a single pixel or that even has a thermal contrast below the noise equivalent temperature difference of the WFOV thermal imager can be detected. Thermal anomalies which have a signal commensurate in amplitude and spatial extent to the object of the search are selected by the system and their coordinates are relayed to the NFOV thermal imager. The NFOV thermal imager zooms into these locations sequentially and presents the image information to the human user who can then either reject or verify that the subject being imaged is the object of the search. |