A B S T R A C TElectrocardiogram (ECG) is a periodic signal reflects the activity of the heart. ECG waveform is an important issue to define the heart function, so it is helpful to recognize the type of heart diseases. ECG graph generates a lot of information that is converted into an electrical signal with standard values of amplitude and duration. The main problem raised in this measurement is the mixing between normal and abnormal, in addition, sometimes, there are overlapping between the P-QRS-T waveform. This research aims to offer an efficient approach to measure all parts of P-QRS-T waveform in order to give a correct decision of heart functionality. The implemented approach including many steps as follows: Preprocessing, baseline process, feature extraction, and diagnosis. The obtained result indicated an adequate recognition rate to verify the heart functionality. The proposed approach depends mainly on the classifier process that based mainly on the extracted ECG waveform features that achieved from exact baseline detection.