ECG signal conditioning by morphological Filtering
- Communications Toolbox
- (Image Processing Toolbox, if you want to use the build-in functions like: imopen, imclose, imerode and imdilate to process the signal)
- ECG_MMF.m: main class of the project.
- Gen_Drift.m, Gen_Noise.m and Gen_Strel.m: generation of structures and artifacts useful for the project.
- erosion.m, dilatation.m, closing.m and opening.m: definition of the morphological operators.
The directory 'dataset' contains a series of ECG signals recovered from the physionet database
The functions 'erosion.m', 'dilatation.m', 'closing.m' and 'opening.m' are used to process the signals. The functions 'GenDrift.m', 'GenStrel.m' are used to generate respectively a baseline drift in the signal and to generate a bunch of Structuring Element to operate on signals.
If you do not want to use the Communication Toolbox with the 'awgn' (Add White Gaussian Noise) function, instead you can adopt the 'GenNoise' function to generate noise.
To built this project I have used the following papers:
- ECG signal conditioning by morphological filtering
- Impulsive noise suppression and background normalization of electrocardiogram signals using morphological operators
The pdf file named as 'Progect_Review' consist in a simple presentation of the project.