Application Of Computer –based Examination System
[A CASE STUDY OF ABIA STATE POLYTECHNIC, ABA]

The Dallas country county community college distinct begun considering various biometrics options for online examination in 1999 (Tulloch and Thompson, 1999). Brooks (1997) proposed face recognition as an automatic verification method during examination, in installing digital cameras, so that students photographs becomes part of the school examination records (Thomas, 2001).
Two different types of examination situations can benefit from biometric identification. Biometric systems could be used to authenticate a student’s identify when entering or leaving a large group examination hall or a supervised computerized examination facility. Also biometrics could be used as part of a security system to supervise students while they are writing entrance examination online. Each of the biometric options has specific capabilities that makes each of them better for some situation than for others. Online examination is the most reliable for all the schools examination situations. The objective is to replace proctoring during a school examination with an automated system. Biometrics system to protect access to the internet and world wide web services have been developed, the systems are based on finger prints (Jam and Hong. 1996) hard geometry (Jain, Prabhackar and Ross, 1998) and voice (Boves and Koolwai, 1998).
There are concerns, which makes the school examination applications different from other biometrics applications. The users must login applications benefits from protection by a security system and choose not to defect it. In most cases, the people trying to defect these system are non-users. With online examination, however, it may be to users advantages not to detect the system.
Enrollment should occur before the students has even begun class. Every effort should be made that the enrollment process is well  supervised and verified. The ideal system needs to be capable of providing continuing, transparent, positive identification of the person sitting at the computer keyboard. Ideally, the examination would not begin until the student is identified positively. Then, while each question is being answered, the system would provide continuing periodic positive identification of the students. The students should be able to read in between questions. Only a camera-based system can provide this type of identification. A small video camera, mounted on the top of the monitor could capture the student’s image at certain time intervals without interrupting their work (Richards, 1997), since nearly all video camera installations include a microphone with a sound board, layering face recognition with speaker verification could provide more accuracy in the future without adding to the student’s hardware expenditure.
As cited earlier, in this study, accuracy is also improved in biometric recognition by using multiple image captures for comparison. In an online examination situation many images would be captured at random times. Recognitions of the majority of the others would be used as a declaration of recognition of the examiner.