Advantages and disadvantages of biometric authentication

biometric authentication system face detection

Authentication means a way of verifying the identity of an individual, while biometrics means the use of the biological or behavioral characteristics of humans that are unique. Biometric authentication means using the unique biological characteristics of a person to verify that the user is the one that they claim to be. The user data is first recorded and stored in a database. When a user comes for authentication the system compares the stored data and the real user authentication feature being presented. If the feature matches what is on the system they are authenticated.

The main types of biometric authentication techniques are fingerprint, face recognition, iris, retina, and signature palm among other methods used. Their main advantages are they are secure from data threats, unique, don’t require remembrance, and are simple to verify.

Types of biometric authentication methods

Fingerprint identification

The technology has been used for a long time mostly by security organizations and governments for identification. The fingers prints of individuals don’t change even as they grow old. This makes them a good feature to use because of their permanency and uniqueness. The technology is in use today in mainstream devices like smartphones and laptops. The device comes with scanners that can be used to scan the finger and record data on software for reference.

Retina scan

The retina is the back of the eye where the image is formed. The main feature identified is the blood vessels that are unique to people and also they don’t change. This makes the retina a good part to use in biometric authentication. Retina scan systems are not yet common to ordinary users.

Iris recognition

It is the use of technology that scans and identifies the iris of the eye for authentication. Iris is the internal part of the eye and it is protected from any damage, and it is also unique to all individuals. This makes it suitable for authentication. If the iris data is recorded well it can be very accurate. Just like retinas they are not commonly used.

Face recognition

It involves using face features/morphology to identify the user. It has been used by people to identify each other but it is hard to computerize the identification because of different features that need to be recorded. Face recognition can be used in security companies to identify people who are captured on a CCTV camera or any other camera from the government or company database of face details.

The Facebook company was using face recognition to identify people who are tagged in a given picture. Because of many privacy issues that came with that move they have removed that feature as communicated on their website.

Voice recognition

It is the identification of the tone, sound pitch, and other characteristics of the sound. The sound data capture is difficult to record and again it may change e.g. if the user is sick.

Signature

A person’s signature is not a biological feature but a behavioral biometric one. The signature is the most commonly used biometric modality. They were used even before digital scanners were designed to scan and use biological features. It checks features like angle, and the intensity of the signature is checked.

Other techniques include palm recognition, DNA, and hand geometry

Application areas of biometric system 

Biometric systems have advanced and biome cheap to the level that they are implemented to the mainstream user.  They are no longer used only in areas that require high-security levels like on a military base but they are in our smartphones. 

The latest laptops also come installed with a fingerprint scanner for identification. With the outbreak of Covid19, most learning institutions also implemented the face recognization system where students don’t touch the surface of the gadget. In the automotive industry, cars are using biometrics to identify the owner before they can start with facial recognition. These are among many areas where biometric authentication is used in the day-to-day life of the ordinary user.

Advantages of using biometric systems for authentication

  1. Secure: compared with other authentication methods like passwords it is hard to fake. This makes it a more secure option and hence reliable. The system that uses biometric authentication systems is hard to hack into compared with other types.
  2. Easy and faster verification process: using biometric authentication can eliminate the use of documents such as ID card and other documents which takes a long time to verify. This means the services will be faster and more effective. Again documents can be stolen or misplaced which is not the case for biometric attributes.
  3. No need to remember: biometrics uses physical features that you will always have contrary to other authentication methods which require the user to remember a password and username.
  4. Unique: the features are guaranteed to be unique unlike passwords which can be shared by more than one user, hence simple to hack by guessing.

Disadvantages of using a biometric system

  1. Body parts may be damaged: some features may change due to environmental or work-related factors. For example, a tea-picking farmer’s fingers may be damaged due to frequent use of the finger to pick tea. Those who work in mines or quarry their hands will be affected by the frequent use of tools that end up damaging the hands which makes the system recognize them very hard.
  2. Expensive: most biometric technologies have been used only in area that requires extra security because they are expensive to acquire and implement. Although currently some technologies are being used in the mainstream like fingerprint identification and voice in smartphones. Face recognition is also used in vehicle industries to identify the driver before starting the engine.
  3. Privacy concern: users may not be willing to use the technology for fear that the data may be used to track them down which breach their privacy right. Again how will data collected be stored and will it be shared by who? These questions among other related issues make users shy away from the system that uses biometric authentication.
  4. Characteristics like sound are hard to verify because of the details that need to be captured for them to be recognized.

Criteria to consider when selecting biometric modality.

Biometric modalities are the methods that are used to identify the biological part that can be used for authentication. The biological characteristics that are chosen should meet the following criteria to be used as authentication characteristics. 

  1. Unique: the characteristic should be unique to each user. This means no two people should have the same morphological characteristic or body feature that you choose to use. A fingerprint for example is unique to each individual even if the individuals are twins.
  2. Permanency: the characteristic should be permanent and should not change. The body characteristic to be used is permanent and it is not expected to change even as the user gets old, unless there is an accident and they lose the body part that is used for the system. Physical characteristics should be easy to record for use and authentication without stain.
  3. Be a universal attribute: the body attribute in use should be shared by all users. Every user should have the feature that is being used such as fingerprint, iris, etc.
  4. Resistant to damage: the feature should be resistant to change from all factors that may be influencing it and even if they change it is a very insignificant change that the system cannot detect it.
  5. Secure: the characteristics should not be able to be copied or borrowed from user to user.

Important factors to consider when selecting a biometric system for business.

  1. User-friendly: whichever system is selected should be simple for the user to use without requiring training on how to. It should be friendly for ordinary users and also the system administrators to manage.
  2. Simple to collect data: the system should offer a simple technique to capture and record the biometric data. Whichever biometric modality will be used should be simple to collect the required data. For many users like in a college setup, it would be best to use fingerprint or facial recognition since they are simple to collect from students.
  3. Acceptability by user: the method selected should be acceptable to users of the system both system administrators and ordinary users. The identified biological characteristic identified should be agreeable to most users on the usage.
  4. Compatibility with existing systems: legacy systems have user data that will be required to be moved to the new system during system changeover. Ensure that the data from the old system are compatible with the new system.
  5. Accuracy: the system should be able to capture accurate data that can be relied upon. If the system is not accurate then it will deny access to even the authentic user of the system.
  6. Business and government policy: to be on the right side of the law both ensure your ICT policy and government policies allows to collection and use of biometric data and how it should be stored. Biometric data collection and use is a major issue when it comes to data privacy. 
  7. Cleanliness of physical data collection hardware: the use of fingerprints and palms are contact types of a biometric system. With the spread of Covid19 and also for hygiene reasons the company may choose to use contactless systems such as face recognization and iris scans.
  8. Initial installation and maintenance cost: cost of acquiring and installation of the system should be considered. The cost will be determined by other factors such as the modality that will be used, the scanner, the hardware required, etc. After installation, the cost of maintaining the system should also be factored in.
  9. Best modality: the biological characteristic to be used should be considered. Fingerprint and facial recognition are the most common types that are used. Choose a simple to collect data that is cheap and user-friendly method.