Simple method to reduce the memory effect in microbolometer cameras

Motion blur is frequently observed in images captured by infrared (IR) microbolometer cameras. This is due to the important memory effect in this type of sensor. Actually, it corrupts the response not only in the presence of relative motion between camera and target object, but also in the event of rapid variations in the temperature of a static target. From a thermal point of view, a microbolometer sensor behaves like a first-order dynamic system. This is why the sensor’s temperature, hence the output signal, retains traces of all past IR radiation events. This memory of past events is fading exponentially with a timescale corresponding to the sensor’s so-called thermal time constant. We propose an image processing method that eliminates, at least partially, the memory effect of microbolometer cameras. It considers each pixel individually and removes from the current signal the influence of incident radiation history that is older than the pixel’s previous reading. Specifically, in any image, the memory of IR radiation events that took place before the previous image can be completely erased. Motion blur is therefore considerably reduced. Thanks to this patented procedure, the microbolometer camera can be considered as having evolved from a time-constant sensor (i.e., with infinite support) to an integration-time sensor (i.e., with finite support), much like cameras equipped with cooled photon detectors. This specific integration time corresponds to the delay between two readout cycles. Given this, it is possible to further reduce motion blur by increasing the camera’s readout frequency. The proposed method does not require complex deconvolution procedures, is independent on the direction of relative motion, is easy to implement, and can operate in real time.