<p>Monitoring heart rates has become a necessity for monitoring cardiovascular health and early detection of anomalies. Missing detection leads to failure in the early diagnosis of arrhythmias, tachycardias, or bradycardias. If not diagnosed and treated, these carry serious health consequences, including stroke and sudden cardiac death. The method proposed in this work meets the requirement of a reliable heart rate detection system, which uses an advanced signal processing and machine learning framework method. The methodology here would include extracting frames from facial video recordings that have been self-collected from real-time individuals into the green channel that has been processed for PPG signals and employing advanced techniques such as Wavelet Transform, Empirical Mode Decomposition (EMD) and peak detection to estimate heart rate non-invasively in real conditions. A major feature of it would be adaptiveness to different environments or user conditions, so it could be expected to perform robustly in many similar contexts. This discovery, as a result of its scalable approach for continuous and discreet heart rate monitoring, would likely work very well in wearables, improving the whole area's management of cardiovascular health. The conclusions presented in this research provide very timely perspectives and strategically position new pathways in the future toward developing health monitoring systems.</p>