Automatic Face Tracking and Attendance System Using Machine Learning Techniques
Ranjan Rufus Rozario S, Manjunatha Hiremath · International journal of tomography and simulation · 2018
The human face is the most significant part of our identity. It plays a very important role in daily communication, interaction and other routine habits in our lives. The main advantage of facial detection over many other biometric systems is the uniqueness of the subject. The human face has a very high rate of variability and hence this makes its detection and/or tracking a difficult problem in computer vision. Hence, face detection and tracking algorithms are necessary for human computer interaction, intrusion detection (Security), etc. Extensive research is being done in this field. In this research work, the face detection and recognition is experimented with improved Viola Jones and FLD algorithms. Further, this detection is extended to tracking in real-time video using an improved KLT algorithm. The experimental results shows 98.63% recognition and tracking accuracy. To show the real time performance of these enhanced algorithms, the automatic attendance system is considered as an application. Attendance is provided to the students present in the image captured and processed using the proposed algorithms. Since the use of machine learning has been widely accepted, the proposed recognition algorithm uses machine learning techniques to train the system and recognise the students present in the image. The proposed approach uses the technique of increasing the efficiency at every level of the system. Efficiency is increased at the face detection, face recognition and the face tracking phase of the system. Hence, this substantially increases the overall performance of the system.