EYE BLINK DETECTION-TRACKING AND ANALYSING BLINKING PATTERNS

EYE BLINK DETECTION-TRACKING AND ANALYSING BLINKING PATTERNS

K.Siddartha, SINDHU REDDY KASIREDDY, B.Sai Varun Reddy, N.Sindhu, A.Harish, B.Siva Prasad, Preeti

School of Engineering

Computer Science- AI&ML

Malla Reddy University,India.

ABSTRACT

Eye blink detection plays a significant role in various fields, such as human-computer interaction, driver drowsiness detection, and health monitoring. This paper presents a Python-based approach for real-time eye blink detection using computer vision techniques.The proposed method utilizes libraries such as NumPy, OpenCV (cv2), and dlib to capture a live video stream and detect eye blinks. Facial landmarks obtained from the dlib library are used to locate the eyes and calculate the eye aspect ratio (EAR) for each eye. The EAR is a reliable indicator of eye openness, and by monitoring its value, eye blink events can be detected.The algorithm involves several steps: facial region detection using a frontal face detector, extraction of eye landmarks, computation of the EAR, and determining if a blink has occurred based on a predefined threshold and consecutive frames. The number of eye blinks is continuously updated and displayed in real-time.Experiments were conducted using a webcam, and the proposed method demonstrated accurate and efficient eye blink detection. The system successfully identified eye blinks in various scenarios and environments, making it suitable for applications that require real-time eye monitoring.The results show that the developed Python-based eye blink detection system provides a robust and cost-effective solution for eye blink analysis. Its implementation using widely available libraries simplifies the setup and usage, making it accessible to researchers and developers. The system can be further integrated into applications such as driver monitoring systems, human-computer interfaces, and health monitoring devices, enhancing their functionality and improving user safety.