How to Select a Suitable Biometric System for Your Project

Biometrics has gained traction in multiple large-scale deployments around the world. With the rise of security thefts nowadays, governments and organizations worldwide are embracing this technology for vast applications ranging from government sectors such as national ID, border control, law enforcement to commercial ones such as education, banking, and hospital management. Some notable projects include US Exit-Entry Program, Indian National ID (Aahar project), and European Smart Borders. Such far-reaching projects have brought indisputable benefits to humankind, and biometrics has been gaining more and more public acceptance.

Until now, there has been many types of biometrics used around for decades including fingerprint recognition, facial recognition, iris recognition. As a decision maker for a biometric project, your major concern would be what type of biometric modalities is optimal for the project. Choosing an appropriate biometric modality is of utmost importance for the success of the project, and the fact is there is no one-size-fits-all solution.

Hình ảnh có liên quan

In reality, many factors which can have a potential impact on the performance of biometric modality need to be taken into account during the procurement process. Firstly, it is worth noting that the performance and cost of the system greatly depend on the environment specifically working conditions of people who will use the system and the way they operate it. A typical example of the environment factor is lighting condition affecting greatly the accuracy of the system such as some types of facial, or iris scanners might not perform well under the poor lighting condition or direct sunlight. In addition to budget and environment issue, understanding the pros and cons of each biometric modality can assist you in making a prudent and wise procurement decision.

In terms of accuracy, DNA is seen as the most accurate biometric modality currently as the chance of two individuals sharing the same DNA profile is less than one in a 100 billion with 26 different bands studied; however with its unique features, it can only be used in specific circumstances. Meanwhile, fingerprint recognition is the most widely used biometric trait thanks to its long history of user acceptance, numerous sources (10 fingers) available for collection, low cost and high rate of accuracy. Recently facial recognition is being deployed more by immigration and border control agencies across worldwide airports thanks to its hygiene, high usability and cost-effectiveness despite the fact that it does not have a high rate of accuracy. Voice recognition is also making inroads into more real-life applications especially banking sectors thanks to its ability of conducting a remote authentication and rich availability of devices for collecting speech samples (e.g., telephone network and computer microphones) and its ease of integration.

Notably, the most reliable and fastest biometric method equipped with a very good positive identification against a large database is iris recognition which is seeing in high-security applications and large-scale deployment such as national ID, non-profit, military, banking and hospital thanks to its utmost accuracy. Ten years ago, widespread adoption of iris recognition has been hindered by the high cost of devices; yet the mainstream usage of this technology in Aahaar project, the world largest biometric project as of now, and its superior accuracy over fingerprint recognition proven in the field tests by Aahaar and UNHCR project together with the increasingly cost reduction has fueled its public acceptance and widespread application in both government and commercial areas in the past few years.

Last but not least, the scale of the project also significantly influences the decision of choosing biometric modalities as the accuracy, usability, speed  might become barriers to an efficient implementation of any project with huge database size such as national ID, border control. In the past years, for large-scale projects, the usage of multimodal biometric systems has been achieving prominence as collecting more than one piece of biometric data from individuals enables to make identification more accurate. Not mention to that; the combination of different modalities is to improve the recognition effectiveness and effectively deal with a variety of issues such as noisy data, intra-class variations, limited degrees of freedom, non-universality, spoof attacks, and unacceptable error rates which may be caused by unimodal biometric systems.

Taking an example of the national ID project in India, with the huge population of laborers, incorporating both fingerprint and iris biometrics is the optimal solution which contributes to the phenomenal success of the project with almost 1.3 billion Indian people having registered by both fingerprint and iris as of September 2018. Similarly, United Nations of Refugee Agency and World Food Program have collected facial, fingerprint and iris data to authenticate refugees around the world.

Briefly speaking, besides to the cost and environmental issues, other critical factors need to be considered consisting of accuracy, usability, user acceptance and hygiene, and you need to evaluate the plus and minus sides of each biometric modality to choose the most suitable solution for their project.