Accurately differentiating between ventricular fibrillation (VF) and ventricular tachycardia (VT) episodes is crucial in preventing potentially fatal missed interpretations that could lead to needless shock to the patients, resulting in damaging the heart. Apart from accurately classifying between VT and VF, the predetermination of the onset of the ventricular arrhythmias is also important in order to allow for more efficient monitoring of patients and can potentially save one’s life. Thus, this research intends to focus on developing a system called Classification and Onset Determination System (CODS) that is able to classify, track and monitor ventricular arrhythmias by using a method called Second Order Dynamic Binary Decomposition (SOD-BD) technique. Two significant characteristics (the natural frequency and the input parameter) were extracted from Electrocardiogram (ECG) signals that are provided by Physiobank database and analyzed to find the significant differences for each ventricular arrhythmia types and classify the ECGs accordingly (N, VT and VF). The outcome from these ECG extractions was also used to locate the onset of ventricular arrhythmia that is useful to predict the occurrence of the heart abnormalities. All the ECGs analysis, parameters extraction, classification techniques, and the CODS are developed using LabVIEW software.