Abstarct: Physiological parameters such as heart rate are among the most important indicators of an individual's health and are typically measured using contact-baxsed methods such as electrocardiography and contact-baxsed photoplethysmography. These methods face challenges, including the need for direct contact with the skin and sensitivity to movement, which can cause issues in specific situations, such as monitoring infants in intensive care units. In this study, we propose a hybrid approach combining traditional methods and deep learning to extract the rPPG signal from facial videos, addressing two common artifacts: lighting changes and video quality, which negatively impact the accuracy of these signals. The proposed method employs two distinct face detection approaches: the S3FD neural network-baxsed face detector and the landmark-baxsed face detection method. For facial skin detection, we use a combination of the RGB, HSV, and YCbCr color models. After detecting the face and skin, RGB signals are extracted and normalized, followed by filtering using the Savitzky-Golay filter. The extracted signals are then converted to PPG signals using CHROM and POS methods, and their power spectral density is estimated using the Welch method. Finally, a one-dimensional convolutional neural network (CNN) is used to train and evaluate the performance of the proposed method. The results of this study demonstrate that the proposed method effectively reduces the negative effects of movement and lighting changes by using landmark-baxsed face detection and enhancing video quality via Eulerian magnification, allowing for more accurate extraction of rPPG signals in real-world conditions. For instance, in the UBFC-rPPG dataset, the proposed method outperforms other methods, including ICA, POS, CHROM, and RSVR, with an AAE of 1.67 ± 0.40 and an ARE of 1.71 ± 0.35. Additionally, in the COHFACE dataset, the proposed method demonstrates superiority over methods such as DeppPhys, HR-CNN, and PhysNet, with AAE values of 2.32 and RMSE of 7.12. These achievements have the potential for broad applications in various fields, including medicine and remote healthcare technologies.