The Ministry of Information Technology and Telecommunications Pakistan is working on the development of its own facial recognition system which it will launch soon to provide strong video surveillance, it was revealed on Wednesday by official resources at Ignite (AKA—National ICT R&D Fund).

The system would not only add value to the existing security standards in Pakistan and international research but also a prominent contribution to the society. Once the system is developed it would also be commercialized and available as stand-alone FPGA-based prototype solution, startup companies could take the system and market them in video surveillance category—The News reported.

The project—Design and Development of an FPGA-based Multi-Scale Face Recognition System will cost roughly Rs.13.84 million which is being developed by PAF-Karachi Institute of Economics and Technology.

The sources said video surveillance along with facial recognition is one of the best security measures and surveillance. Increasing security threats and the need for invulnerable security systems are becoming inevitable.

A recent conference in Islamabad organized by the Pakistan Information Security Association (PISA) called upon removing security vulnerabilities from the digital landscape of the country. This system could also play a significant role in this effort. Traditional security measures can be easily fooled by hackers, smart cards, keys, and passwords are not as secure as facial recognition. Among the other available biometrics, such as speech, iris, fingerprints, hand geometry and gait, the face seems to be the most natural choice. It is non-intrusive, requires a minimum of user cooperation and is cheap to implement—the sources said.

How The Facial Recognition System Works

The system that is being developed focuses on high-speed FPGA-based multi-scale face recognition system which uses Linear Binary Pattern (LBP). The most important element in the system is algorithm design which will be mapped in Hardware.

LBP extracts texture information from the face, Wavelet decomposition of these features result in sub-bands encompassing low and high-frequency components, and the sub-bands contain useful information for classification, the sub-bands have various categories, some are more efficient than others coupled with intelligent selection of these sub-bands would likely to increase the performance of face recognition system. Therefore, the identification of discriminant sub-bands for efficient and robust face recognition is the main objective of the system to achieve.

According to the sources the system would be soon launched in Pakistan adding an extra layer of security for the country.