Temporal statistics de-ghosting for adaptive non-uniformity correction in infrared focal plane arrays

The LMS-based adaptive non-uniformity correction (NUC) technique, known in the literature as Scribner´s algorithm, is an efficient method to mitigate the presence of fixed pattern noise in video sequences acquired by thermal cameras. Unfortunately, ghosting artefacts can be generated from the process of NUC depending on the edges that characterise the scene inside the sensor´s field of view. Introduced is a de-ghosting technique based on the computation of the statistics of the error signal employed in the Scribner´s algorithm. Such a technique is characterised by a small computational load which is the most important requirement for real-time applications. Experimental results demonstrate its good performance and de-ghosting capabilities.