Comparing Fingerprint and Facial Recognition Technologies for Employees Tracking

“Biometrics” have been a buzz word in the technological circle in recent years, as it is becoming increasingly accessible and accepted by the market and industries. Biometrics are the unique traits which identify each and every one of us as individuals, specific by our own distinct features that are not replicated by another. Traits such as facial and body features, voice, mannerisms, are just some ways we, as human beings, visually identify one another. There are other traits, which are more minute and difficult to detect, such as fingerprints, DNA, vein or iris patterns.

Ever since mankind had known that biometrics are unique properties to each and every human being, we have tried to apply it as a technological means to efficiently aid us in real world scenarios. Since BC 200, ancient Chinese records from the Qin dynasty include details about using handprints as evidence during burglary investigations. In 1891, Juan Vucetich, an Argentine police official, began the world’s first fingerprint files for identification for police work.

article-1 Comparing Fingerprint and Facial Recognition Technologies for Employees Tracking

Today, with the advancement of computing prowess and technological breakthroughs, the implementation of real-world usage of biometrics came true through biometric technologies. Since the early 2000s, fingerprint technologies are applied in the security and human resource sectors, in specific use-cases such as high security access control and manpower tracking. During that period of time, Intercorp was the first company in Singapore to offer such services.

Fingerprint recognition was widely accepted by the market and its adoption rate picked up swiftly, becoming the de-facto biometric trait. Its adoption success could be attributed to it’s straight forward concept that every human being’s fingerprint patterns are visible and unique, and the trust of it being a thoroughly proven method as it had been used in police forensic work for many decades.

However fingerprint recognition technology is not without its flaws and restrictions. Due to its methodology of comparing ridges, valleys, arches, loops and whorls, its accuracy can be compromised when these attributes are contaminated. If one were to wash his or her hands and try to scan on the fingerprint reader, possible remaining water residues would affect the accuracy of identification. Genetic conditions such as sweaty palms would fall in such similar categorisation as well.

article-2 Comparing Fingerprint and Facial Recognition Technologies for Employees Tracking

Being a fragile and physical attribute, fingerprints can be worn off too. Work such as labouring and typing wears down the fingerprint ridges and affects the smoothness of the skin. This make the fingerprint very hard to read, which could explain why prisoners have been known to rub their hands against rough walls of prison cells to try to wear away the ridges. Hence the work demand, environment, age and genetic conditions of a person determine the rate of identification accuracy.

In 2003, an outbreak of SARS affecting 37 countries and resulting in 774 deaths raised questions about the co-relations between epidemic and close human interactions. The application of contact-based biometrics, used by large numbers of users, was raised as one of the key concerns. Our company, Intercorp, similarly faced such questions by our clients as we treated the matter seriously and stepped up sterilisation as a standard maintenance process. However it hit us hard then to rethink out of the box and to find a more viable solution through a non-invasive, contact-less solution.

After years of working with partners and fine-tuning identification algorithms, in 2009, facial recognition technologies finally became a viable solution. Instead of relying on simple methodologies of image comparison, Intercorp facial recognition captures the subject’s facial dimensions through active scanning via dual lens cameras for depth analysis and unique facial points identification and a face would typically yield about 6,000 unique feature points which can be computed for match analysis.

article-3 Comparing Fingerprint and Facial Recognition Technologies for Employees Tracking

Facial authentication is faster than fingerprint recognition as well, with 99.999% accuracy rate and an authentication speed of less than 1 second, ensuring smooth usage even for heavy human traffic flow. One of our busiest project, with more than 15,000 employees on-site, facial recognition has proven to be easier to deploy without issues. Being a contactless solution, it requires less maintenance as well, without components susceptible to daily wear and tear.

In the corporate world, with varying workplace environments, work trades and processes, the technologies employed by solutions have to be robust and indifferent to the environment they are deployed in. If accuracy or operations can be adversely affected by physical constraints, such as water residues in the fingerprint case study, then such implementations are short-changed from the very beginning.

Hygienic, accurate, reliable and easier to maintain, its no wonder that facial recognition technologies is currently widely appreciated by users and the general public. Eventually it will definitely displace fingerprint as the de-facto biometric trait.