A Thermal IR Processing Technique for Detecting and Tracking Objects in Motion

The detection and tracking of thermal infrared targets is usually associated with the military. However, there are a number of interesting applications in the civilian sector. We detail a digital imaging process technique for the detection and tracking of up to 3500 moving objects per minute in a complex background. Uncalibrated video-rate thermal imagery from a stationary FPA micro-bolometric IR imager is captured on 8- bit digital media. Frames are then differenced with an adaptive temporal background to remove stationary clutter, and thresholded to select pixels at the tail ends of the distribution of differenced pixels. Moving objects then appear as pixel clusters in the image frame. If similar-sized pixel clusters from two successive frames are found within a selectable error tolerance, then a new track will be established. The location pairs define a motion vector that is updated every frame and is used to predict the next position. This process is iterated frame-by-frame, generating an output file of individual object tracks. An example is illustrated using bats emerging from a roost site.