Optimized Principal Component Analysis Algorithm Based Facial Recognition System with Liveness Detection for IoT Applications

Ali Akbar Siddique, Muhammad Tayyab, Muhammad Tahir Qadri, Adeel Sarwar, Sidra Ahmed


Driven by commercial applications and daily basis security requirements, research, and analysis on facial recognition algorithms from video sources has become an area of growing interest among the researchers in recent times. In this paper, an algorithm is proposed based on facial recognition system incorporated with liveness detection that employs Principal Component Analysis (PCA) to extract and reduce the dimensions of the facial features and Support Vector Machine (SVM) is utilized to train these features to assure the physical presence of a person being detected. The system requires a real time video feed through which it extracts an individual frame that contains a detected face of a person using Viola Jones algorithm. Facial recognition process is followed by an eye blinking algorithm through which the system will identity the actual presence of a person on secure environment. The main objective of the proposed algorithm is to provide secure access to the authorized person and upload the information regarding his activity on the cloud server. Thus, the proposed system can be used to reduce the manual work involved in terms of operation and at the same time increase its efficiency. The overall efficiency of facial recognition system was found to be 90.72% and the face liveness was found to be 94.07% for the dataset used.

Full Text:



  • There are currently no refbacks.