Fast algorithm for face detection and tracking??

<font color=#550080>The conventional personal identification techniques like passwords, keys, barcodes and smartcards have a major drawback: they don�t check who is entering or holding the information. They only check whether the correct information is presented to the system. Such</font> systems can be easily deceived because any person who has the ID card or

anyone who knows the password can easily claim the identity of that person. Also there are some other drawbacks for such systems: the person has to remember the password or he has to carry the ID card. In barcode system, line of sight reading is required. Biometrics based personal identification systems eliminates most of these drawbacks. Biometrics is the automated recognition of individuals based on their behavioural and biological characteristics. Fingerprint, hand geometry, face, iris scan, retinal scan, signature, gait, voice are some of the well known biometric characteristics. In biometric based attendance systems, subject is being identified using any of biometric techniques Biometric systems are more reliable because biometric characteristics cannot be easily stolen, duplicated or lost and also the user does not have to memorize the password or he does not have to carry some ID cards. But each of these biometric systems has certain disadvantages also. The finger print and hand geometry based systems fail to identify the individuals if the finger or hand is injured or dirty.Retinal and iris scan are very much susceptible to diseases that change the characteristics of the eye. For obtaining the retina scan, laser light must be directed through the cornea of the eye. Iris based systems need a specialized camera which is very expensive and also the photo should be taken very close to the subject. Voice based systems might not work properly if the voice of the person changes due to flu or throat infection. <font color=#5566B3></font>