A morphology based algorithm for baseline wander elimination in ECG records

Morphological filtering techniques are considered in an attempt to eliminate baseline wander in ECG records. Using the sequence processing morphological filters that utilize the basic transformations of mathematical morphology, the authors have processed ECG records off-line. As the results clearly indicate, the technique selectively removes the baseline variation without introducing any distortion that the hardware or software linear filtering may introduce. The promising results obtained also encourage the use of morphological filtering techniques to solve other filtering and enhancement problems related to ECG signal. Besides their power in removing the baseline wander, the morphological filters are very easy to implement in general.<<ETX>>