Of the various biometric identification methods, face recognition is one of the most flexible, working even when the subject is unaware of being scanned. It also shows promise as a way to search through masses of people who spent only seconds in front of a "scanner" - that is, an ordinary digital camera.
Face recognition systems work by systematically analyzing specific features that are common to everyone's face - the distance between the eyes, width of the nose, position of cheekbones, jaw line, chin and so forth. These numerical quantities are then combined in a single code that uniquely identifies each person.
Fingerprint identification
Fingerprints remain constant throughout life. In over 140 years of fingerprint comparison worldwide, no two fingerprints have ever been found to be alike, not even those of identical twins. Good fingerprint scanners have been installed in PDAs like the iPaq Pocket PC; so scanner technology is also easy. Might not work in industrial applications since it requires clean hands.
Fingerprint identification involves comparing the pattern of ridges and furrows on the fingertips, as well as the minutiae points (ridge characteristics that occur when a ridge splits into two, or ends) of a specimen print with a database of prints on file.
Hand geometry biometrics
Hand geometry readers work in harsh environments, do not require clean conditions, and forms a very small dataset. It is not regarded as an intrusive kind of test. It is often the authentication method of choice in industrial environments.
Retina scan
There is no known way to replicate a retina. As far as anyone knows, the pattern of the blood vessels at the back of the eye is unique and stays the same for a lifetime. However, it requires about 15 seconds of careful concentration to take a good scan. Retina scan remains a standard in military and government installations.
Iris scan
Like a retina scan, an iris scan also provides unique biometric data that is very difficult to duplicate and remains the same for a lifetime. The scan is similarly difficult to make (may be difficult for children or the infirm). However, there are ways of encoding the iris scan biometric data in a way that it can be carried around securely in a "barcode" format. (See the SF in the News article Biometric Identification Finally Gets Started for some detailed information about how to perform an iris scan.)
Signature
A signature is another example of biometric data that is easy to gather and is not physically intrusive. Digitized signatures are sometimes used, but usually have insufficient resolution to ensure authentication.
Voice analysis
Like face recognition, voice biometrics provide a way to authenticate identity without the subject's knowledge. It is easier to fake (using a tape recording); it is not possible to fool an analyst by imitating another person's voice.
You cannot answer this question directly, because you need to define "best" first, best in what? accuracy, speed, cost, user friendly, availability, etc.
However, there is no perfect biometric, each has its own pros and cons, and therefore it all depends on the system requirements and needs. for example the most accurate is Ires, but its imaging device costs high, on the other hand, keystroke is the cheapest, but the least accurate.
Dr. Mushtaq Ahmad gave above a very good overview of typical Biometric Methods (BM). In general, BM are successfully used at the level of an enterprise to authenticate employees at the door, local terminal, and at the desktop / laptop computer, and alike. The usage of BM for consumer level remote authentication in e- and m-commerce is currently quite limited, for the following reasons:
-Costly integration with hardware and network software authentication protocols
-Costly maintenance and online support
-Cost and security problems with mass BM credentials setup, storage, and reset
-Various values of false negative and false positive errors given by different vendors
-Static BM credentials vulnerability to numerous attack vectors on the front- and back-end
-Reluctance of many consumers to use BM due to cultural and religious concerns
-BM technology allows only user-to-server, but not server-to-user authentication, which requires another technology to enable a mutual authentication protocol, leading to increased cost and complexity
Due to those above and other obstacles, BM is sometimes used only in two-factor remote authentication systems as a fist factor of a user to desktop / laptop / mobile device authentication (local level of user-to-device authentication), and the second factor is based on a user entering a one-time code delivered from a remote server through for example an email or SMS message.
However, during the last decade new BM came into being – so called Behavioral BM (BBM). It utilizes two separate and different patterns:
-First pattern is based on information collected by marketing analyses servers sitting at the back offices of banks, credit card companies and alike to gather peoples’ social behavior (how often and what, where, and how people buy something, or take particular services; what brands of goods and services people used to prefer, etc.). Then, if real time monitoring unveils that a consumer is not acting along its pattern of behaviors – the flag is raised and the transaction is postponed, until it is clarified. This technique is also known as Intrusion Detection System (IDS) or Intrusion Prevention System (IPS). The clear advantage of these systems is no requirements for a consumer to go through a rigorous authentication process, unless a behavioral deflection from the consumer’s pattern is detected in a real time.
-Second pattern is based on the pattern of a consumer’s movement of a mouse pointer at a remote computer terminal – the ways a consumer draws characters, moves the mouse, etc. follows a certain individually specific and unique bio rhythmical pattern. The level of this BBM security is not clear yet. However, it is an interesting development allowing implementing Continues Users’ Authentication (CUA) during the entire transaction session. This capability is unique and, as far as I know, unavailable by any other authentication method utilizing just front-end information.
The best in what sence? Different methods have different advantages and disadvantages. Iris and fingerprint is good in speed and accuracy. Hand geometry is very bad in accuracy. Retina is old and forgotten. Face is good, but can vary greatly during time (even day/week) because of glasses, hair or makeup.
in Versatility Finger prints are widely used biometrics all over the world. Question to your answer is dependent upon the situation in which you are going to use it...
suppose in intensive security area iris recognition based system will be better ... or in combination with two biometrics you can get more safety...
The combination of iris and fingerprints can be considered as the most significant method for identifying a living person where as DNA sequence and palatal patterns can be most significantly used for identifying a dead person.
I fully agreed with the view that the fingerprint safest and that is still safer to have one or two additional parameters, especially today in the era of the general uncertainty in the world!
Fingerprints can be duplicated (source is FBI) ! Comment about retina being stable for life is also not correct; since, it is vulnerable to many pathological conditions affecting blood vessel crossings and hemorrhages (eg.diabetes as one example) especially as one ages, Additionally congenital diseases affect retina.
The only biometric that is truly reliable is the iris because it consistently is stable throughout life with more degrees of freedom than any other biometric and the physiological dynamics of the pupil make it impossible to duplicate; however without the pupil dynamics iris digital photos are,recently, so sophisticated as to simulate another's iris ! The only pathology affecting iris imaging are corneal opacities and they are very rare compared to retina. There are no ophthalmology textbooks devoted solely to iris because of its stability !
I am the ophthalmologist inventor of the iris biometric (1987 patent and 2013 inductee "Inventors Hall of Fame". My new co-invented 3 biometric (iris,ear,fingerprint) non-contact algorithm pat#9,412,002 (2016) is presently under university investigation by me matching birth-mothers and newborns to replace footprint ID. and, shortly ,upon completion, will be submitted for publication in major AMA journal.
The combination of all of them, gives a high degree of reliability to identify without a doubt a particular person; but:
Of the 5 listed, the most reliable are 2, the fingerprint and the scan of the iris.
Visualization of the retina with its pattern in the shape, diameter and distribution of arterial (A) and venous (V) vessels, and AV junctions; Together with the fovea and the physiological excavation in the emergency of the optic nerve, are not very constant, can be modified by diseases like diabetes mellitus, glaucoma, cataracts and arterial hypertension. Therefore, they are not 100% reliable.
Facial recognition was very reliable in the last century, but with scientific and technological advances, especially plastic and reconstructive surgery, it has greatly diminished its reliability. Although there are methods and instruments of Forensic Medicine and Computer for face recognition (anthropometry and reconfiguration / digital computer simulation of facial features) that are still useful.
The above mentioned in relation to facial recognition, is applicable to the geometry of the hand, in terms of shape and volume (square, rectangular, trapezoidal, oval, etc). A contrary situation would be, for being more effective, if it were the "Geography of the palm of the hand"; That is, lines, grooves, folds, etc.
Finally, if DNA had been mentioned in the list, this would be the most reliable and safe of all described above, since it has a reliability of 99.9%.
regards
Jose Luis
La combinación de todos ellos, da un del alto grado de confiabilidad para identificar sin duda a una persona en particular; pero:
De los 5 enlistados, los más confiables son 2, la huella digital y el escaneo del iris.
La visualización de la retina con su patrón en la forma, diámetro y distribución de vasos arteriales (A) y venosos (V), y los cruces AV; junto con la fóvea y la excavación fisiológica en la emergencia del nervio óptico, no son muy constantes, pueden modificarse por enfermedades como la diabetes mellitus, el glaucoma, las cataratas y la hipertensión arterial. Por tanto, no son 100 % confiables.
El reconocimiento facial era muy confiable en el siglo pasado, pero con el avance científico y tecnológico, en especial la cirugía plástica y reconstructiva, ha o disminuido mucho su confiabilidad. Aunque existen métodos e instrumentos de Medicina Forense y de Informática para reconocimiento facial (antropometría y reconfiguración/simulación digital informática de rasgos faciales) que todavía son de utilidad.
Lo antes dicho en relación al reconocimiento facial, es aplicable para la geometría de la mano, en cuanto a forma y volumen (cuadrada, rectangular, trapezoidal, ovalada, etc). Situación contraria sería, por ser más efectivo, si se tratara de la "Geografía de la palma de la mano"; esto es, líneas, surcos, pliegues, etc.
Finalmente, si en la lista se hubiera mencionado el DNA, éste sería el más confiable y seguro de todos los antes descritos, ya que tiene una confiabilidad del 99.9%.
Thanks for this pertinent question...The combination of 5 or more biometric parameters will enhance the recognition system reliability....but how about the future? the question of cloning humans will make these systems obsolete...so may be we will have to go deeper to pheromones or DNA recognition.....or just like Will Smith in a famous movie implant chips in the body....
You have raised very good question. As our fellow researchers comments /answers have shown bio metric recognition is evolving technology, with new and innovative techniques being used for human bio metric recognition.
Enclosed paper shows 3-D skull based technique
Article Human authentication by matching 3d skull with face image using Scca
Opinions if it is technically possible to verify all five listed parameters, I think it would be much greater certainty than when taking one or two parameters!
The idea is to do multi-feature with fusion for single modalities and another opinion to select more than one bio-metrics models as multi-modal and used the fusion of them by different level of fusion the it will be more secure and you will get the good result depend on which model you select and by which technique you applied
In fact, different factors can play a significant role to choose the best biometric data source, including: the application nature, the level of the security required, cost, etc. For instance, the face trait is considered to be less intrusive, while the iris trait is the most accurate one. Therefore, some people try to merge the advantages of two or more biometric traits into what is called multibiometirc systems.
In the present-days the multi-modal biometrics is commonly used to gain appropriate level of security like José Luis García Vigil and Mouad Ali explained below.
It is hard to predict how this field will be evolving in future. Some interestings research towards the remote and touchless identification are conducted including e.g. gait biometrics.