Category: CSE Project Reports

ATM Security via Facial Recognition:

We can use face recognition technology in ATMs (automatic teller machines) centers instead of pin numbers. This technology is already implemented in Pakistani government.

But not only the uses it can have some problems also. It is easy to manipulate the ATM cards due to some problems. With this technology or by using this technology some issues occurred.

Conclusion

Thus, how the Biometrics is a playing a very important role in the today’s daily life. It has become very common in our daily life now. As I mentioned, Biometrics is very important technology especially in the role of Security. It really helps a lot in the issues of Security with is advanced features and applications . Its getting more and more advanced technology day to day. Not only in securing the data, it also helps in various other fields. We will hope that we will have much more advanced features and applications of the Biometrics in the near future.

Face Recognition: 

Presently, face recognition is measured to be moderately imprecise due to the existence of a collection of inconsistency (from 1.39% to more than 13% EER). Like aging, or basically related to outer environmental situations (facial expressions, illumination, poses, textured background), occurs to people over time suitable to various changes. The concert varies greatly due to this method’s depending on the perspective of applications (static images or videos, by means of or exclusive of uniform conditions, or invariable lightning situations) and on the recording conditions.

Face recognition is not proficient enough at this instant to pact with all-encompassing credentials; but it can be helpful in the environment of substantiation or incomplete entrée organize with constraining possession circumstances (through conscription the back-ground must be uniform and the user must look the camera at a rigid space).  

ACCURACY RATE OF BIOMETRICS:

The circumstance of biometrics, three levels of information combination schemes has been recommended: feature of original level, decision level and matching score level. The principle of biometrics is to examine whether the incorporation of face and palm print biometrics can accomplish superior presentation that may not be achievable using a particular biometric display.

Both Independent Component Analysis (ICA) and Principal Component Analysis (PCA) are measured in this attribute vector synthesis perspective. The outcome of the individual palm print and face is compared with outcome of the shared biometrics. It is established that the performance is extensively enhanced in both cases, particularly in the case of   feature fusion usingICAobtaining hopeful outcome with a “99.17%” recognition accuracy rate via a test set sized of forty people.

 Several additional methods extend the feature drawing out procedure which is based on the PCA have been developed to enlarge its accuracy and diminish its computational cost.  It will acquire the improvement of incomplete variations of the face images.

A method called Two-Dimensional PCA (IMPCA or 2DPCA) that uses a 2-dimensional matrix demonstration for faces instead of traditional 1-dimensional design used by PCA. The face depiction is much lesser than the single necessary for the traditional PCA. So this method works

By means of little dimensional data leading to a more statistical representative covariance matrix.

Gottmukkal and Asari developed the Modular Principal Component Analysis (MPCA) with current improvements. The face images are separated into slighter regions and the PCA approach is applied for apiece of these sub-images in this method.

Face recognition which is based on PCA Technologies are not especially correct when the facial expression and the clarification disagree significantly. Modular PCA and IMPCA are latest techniques that aim to unite the most excellent aspects with these job strategies. So this latest approach is called Modular Image Principal Component Analysis (MIMPCA, for dumpy).

The image demonstration vector obtained by the IMPCA performance is statically advance representative than the imaginative PCA-based techniques.

Principal Component Analysis (PCA) based methods for face have achieved an excellent presentation to distinguish and to characterize the anterior individual faces. On the other hand, its accuracy is enormously affected by face image variations like: head pose, explanation and facial expression. The 2D image metrics should be altered into a 1D vector in PCA-based face recognition techniques.

In this manner, this approach considers the complete image matrix to calculate the basis vector of the latest attribute spaces.  Using the traditional PCA by taking into consideration of an image matrix of size m × n the covariance matrix will be of size m× m as an alternative of m. n × m. n obtained using the traditional PCA. Then, the covariance matrix can be obtained by                                             

Where Ā is the typical image of all the preparation samples and Aj represents the j-th image of the training database. As discussed previously about the MPCA technique which was projected by Gottumukkal and Asari, this approach works as follows: the whole image is separated into lesser regions and the feature extraction is computed for every individual of these regions. So, the restricted ledge will be more representative to the region it covers, see figure 2. Because some variations on the face images do not achieve the complete information of the faces. This method takes benefit of the unchanged regions of the face to advance its accuracy rate. 

Expert image analysis: 

Manual analysis of facial biometric facial appearance (frequently termed facial mapping) is principally an aspect of transgression investigation. The face must be of resembling three-dimensional alignment as the head can shift in a figure of directions. Experts in this pasture have been given that rumors to endow with recognition support, primarily in criminal cases, since 1997.

This facial biometrics employs a process of –

• One-to-one matching:  – Arrangement to recommend the contribution of a exact personage whereby further forms of substantiation must be confirmed all the way through a method of checking suspicious or allegations. 

Facial mapping techniques: 

Even though there are several publications concerning facial image comparisons, numerous methodologies engaged for image comparisons are strained from peer-reviewed and established practices contained by other disciplines, Examples of morphological classifications, three-dimensional analysis, and geometric analysis are specified in the writing and transmit frankly to forensic image comparisons for the relay in a straight line to forensic image comparisons for the purposes of classification.

Face Recognition using biometrics:

Statistical analysis of physical characteristics can be defined by biometrics. For the intention of recognition and authorization the word biometrics refers to the capacity or quantity and comparison of physiological human characteristics and quantifiable physical can be done exclusively within the access control and also in security. Biometrics is motionless relatively new; from an access organize standpoint, and however the use of biometrics is graining position. Because this pattern-recognition system overcomes issues related with approved persons having to hold cards or keys. “Enrollment” is a procedure that is used in biometric systems to confine the proscribed data. When the subject’s biometric allusion data has been composed, it is then stored as a digital outline. Submitted biometric samples are compared with the pattern for the reason of denying admission or conceding, but not stored images or the enrollment samples.

Four practical issues must be measured earlier to selecting the biometric technology for exercise in an information center are: 

• Failure to enroll
• Type 1 error, or false reject
• Type 2 error, or false accept
• Crossover error rate (CER)

Even though the individual criteria that discriminate some biometric factors such as “fingerprint” and retinal minutiae are decades mature, the advances in machinery allowing them to expand into realistic for the entrance power purposes are still comparatively latest.

                                 To access the power systems for information centers or IT conveniences in a dumpy list of the biometric technology is on hand which includes the following: 

• “Facial recognition” dealings only one of its kinds attributes of the appearance, counting exterior skin texture such as geometry, or it can use thermal imaging to record the foremost veins and arteries beneath the skin.
• “Fingerprint recognition” analyzes the patterns that establishes on the tips of the fingers. For more 100years they make use of of the fingerprint patterns as exceptional identifiers about the world. Spawn mature methods to defect or bypass identification, mature methods of identification and authentication are also used to eradicate the potential deduction of the finger of an authorized enrollee and using it to gain the access; various manufacturers are used now to determine pulse and temperature as the part of authentication process.
• “Hand-scan geometry” compares a variety of hand measurements, together with the length, width, thickness and surface area of the hand. This technology is used primarily for access control around for many years. However, it also has been incorporated in many times and also in attendance systems. In many commercially presented units temperature sensors have been added to this technology.
• “Other biometric technology” includes several new biometric signatures that can be used to be in command of admission to a secure site or room, such as handwriting technology, DNA, signature, voice recognition, keystroke dynamics and signature dynamics. 

                      Face biometrics is regarded as a smaller amount invasive than additional methods such as fingerprint analysis, iris scans, palm morphologies, which commonly necessitate co-operation from the theme and a responsiveness of the course of action being undertaken.

Fingerprint processing

Fingerprints are not compared and also not stored in the database as bitmap images. Fingerprint matching techniques can be placed into two categories: minutiae-based and correlation based. Minutiae points are first found and then mapped to the relative index of the finger in minutiae-based technique.           

Loop:

The most common type of fingerprint patterns is loop and it account almost 65% of the matches and is as shown below.        

Arch:

The arch pattern is a more open curve than the loop. There are two different models of these particular arch patterns: they are the “plain arch and tended arch patterns” 

Whorl: 

“This particular pattern occurs almost 30% fingerprints and are defined by atleast one ridge that composes a complete circle and is as shown below” 

It is not always sure that fingerprint ridges are straight and continuous 

always. They can be broken, forked and changed directionally. The points where the ridges end or broken are called minutiae points and these points are the main key for unique identification. Ridge endings and ridge bifurcations are the two most important types of ridge points and are as shown below                  

The most important factor the decision speed mainly depends on  few items like the level of security implemented and answers that were delivered in the sense negative answers will affect the speed than the positive answers any more. The speed and the accuracy algorithms are no way dependent on each other. The minutiae found on many fingerprint samples are stored for future comparison studies. Encoding and other compression techniques are also applied on minutiae samples found. In general the typical and traditional formal size of these particular master templates would be around  between 24 bytes and one kilobyte. A very huge set of these desired data is stored in the fingerprints captured. Due to the heavy level of data that is present in any single fingerprint the probability of resulting a wrong match is always less and can be reduced to a very small fraction. These fingerprint technologies can also be targeted and more better ways can be implemented to achieve these targets on a very large databases.

To obtain the accurate results from the fingerprints, there are some pre-processor techniques and steps to be followed and few of them are as below

Evaluating Biometrics

Bio-metric industry has tremendous growth since 3 to 5 years. There was a call for more bio-metric application after the collapse of twin towers on sep 11^th. Those, however, seem to be reluctant to invest in a technology that is still emerging and therefore features a bewildering diversity of products and companies. As a recent article in the Washington Business Journal brought it to the point, “bio-metrics is the next evolution in security, but it’s a very fragmented industry, with lots of companies trying different things” 

Metrics and performance criteria 

It’s always challenging to any bio-metric or traditional authentication systems to provide more secure authentication, identification and verification process. There are lot of mechanisms by which the performance of any system is monitored. All these values should be tracked in a finniest manner irrespective of the methodologies used while capturing the results.

All most all the bio-metric tools uses few buzz words to identify the measuring the factors of the performance. For example, a renowned vendor of fingerprint sensors claims on his website that his “sensor achieved 0.0% false acceptance and 0.0% false rejection rates”. Like this there are lot of attractive and good looking technical standards and measurements are set to measure the performance of the bio-metric systems. There are many metric that we used to measure the performance of the bio-metric system and three among them are as listed below 

False Match Rates (FMR) 

False Match Rate is the probability that the respective system will match the current user verification template with another user verification and enrollment template. It can be understood as the likelihood of an impostor being recognized as a legitimate user. This is the most frequently found error rate and can be handled very critically with any knowledge on the error rates. 

False Non Match Rate (FNMR) 

False Non Match Rate is the probability that a particular user verification template is not matched with the same users enrollment template. Heavy  false match rates could always lead to low or zero productivity or user frustration and may be not that critical as False Match Rate. 

Failure-to-Enroll Rate (FTE) 

Failure-to-Enroll Rate is the probability that the system will not

be able to extract distinctive, consistent and replaceable characteristics from the sample presented during the enrollment process, i.e. this is the likelihood of the system being unable to create an enrollment template for a new user. The affect of this could in the behavioral side rather than physical side. 

All the above Failure rates will impact the system in one or more ways and this affect is depend on the number of iterations worked on the failure rates. The most predominant among all the three key failures are are False Non Match Rate and Failure-to-Enroll. These two plays a very important role in determining the performance of the bio-metric system. In the ideal cases at maximum only three attempts should be done for any metrics and it’s affect operations.

Facial recognition 

Facial recognition is one of the widely and most commonly used biometric application and this can also considered as the natural means of identification technique that were in the market. Any high quality camera with capability capture the images with a very high resolution can be used to capture the images of human faces and also a scanned copy of the image can also be used for this facial recognition process. More accurate results can be produced by the better image that is captured either by a camera or we can use scan copy. Gray-scale information is the primary input for facial recognition systems.

The colors of the final image captured are only taken in to consideration for identifying the face in the entire picture that was captured. Quality of the camera plays a very important role in the lightening conditions. A poor light condition does not make sense to identify an individual facial features and thus a proper steps need to be taken to avoid the poor camera light if there exists any thing. Now-a-days the most advanced infrared cameras can be used for these facial recognition systems. 

For a better picture quality the desired user should be at a pre-defined distance from the particular camera, so that the facial recognition system can get an extra quality input for it’s internal processing. Locating the face with in the entire image would be the very first task of any facial recognition systems and the respective facial characteristics are captured. After locating the face in the image, the system focus on eyes of the face as shown in the below image

Facial metrics mainly depends on measurement of specific features of face like eyes, nose and mouth and also distance between them is also considered. Now-a-days a more sophisticated method for recognizing the face has been developed and widely used since three years. As per this method the faces are categorized based on the angle of fit and with a “fixed set of 150 master eigenfaces. In this process every face is assigned a degree of fit to each of the 150 master eigenfaces and only the 40 template eigenfaces with the highest degree of fit are necessary to reconstruct the face with the accuracy of 99%.” 

A computer software will take care of the image processing and calculates the degree of fit. A lot of intelligent computing power is required to process this technique and assembling a stand alone face recognition system may not serve the purpose. There were lot of efforts now a days to build a sophisticated and chip based embedded system that can be used for facial recognition system which holds a special instruction set.

The accuracy of the computer software is not that ideal and has lot of disadvantages that can be discussed and still there is a lot of potential scope of improvising the algorithms being used as of now. The current software in place may not recognize the face or may identify the face at a wrong place. If the eyes are positioned at the exact place then much better results can be achieved.

The face recognition can be re-sizeable up to an extent like 150X100 points. This normalized shape of the face is called canonical image and the size of this template can be in the range of 3 to 5 KB and is as shown below

Retina Scan 

Retina scan can be considered as one of the important biometric scanning methodologies. This particular process of scanning the retina is mainly based on the vessel pattern of the blood that resides inside the retina of any eye. “The first retinal scanning systems were launched by EyeDentify in 1985”. Infra red light source influences the retina apart from the normal visualization. Blood vessels present inside the retina of eye absorbs the infrared radiations very quickly and the tissue is also affected by this absorption. The retina blood vessel pattern is captured as an image and special characteristic points are analyzed. Below image shows a typical retina patterns 

Intrusiveness can be considered as the main drawback of the retina scanning process. The way the retina scan was obtained can be considered as not safe and is much dangerous when suggested. The cornea of eye is directly affected by the laser light during the retina scan and this process is not that easy to execute and requires a lot of skilled operations and instruction sequences. The volume of data produced by retinal scan is almost same as the fingerprint scan process and thus this scan process can be used both for identification and verification.

In the practice, however, the retina scanning is used mostly for verification. The size of the eye signature template is 96 bytes. Retina scan produces much more accurate results 

Because of it’s less user friendliness and expensive retina scan is not being used widely and is used rarely. This retina scan can be launched and implement in cases where there is a huge scope for security and user authentication. Retina scan systems are used in many U.S. prisons to verify the prisoners before they are released. Below figure shows a typical retinal scanner system that was used commonly

Why Use Biometrics? 

There are lot of advantage in using the biometric based applications. These are used to identify and authenticate many people for most secure gateways. It identifies a persona in many unsafe and preliminary conditions and recognizes you as you. Many important object like tokens, Id cards, ATM cards can be lost or being theft. Most secured identities like the passwords which can be easily forgotten and even in some cases they can be loosed or stolen across the network by different types of hackers around the internet.

           Now-a-days the most high end e-commerce and electronic world is making the end users to remember their personal identification proofs like passwords. These high secured requirements are also posing to create multiple passwords and authentication metrics like ATM pins, Personal authentication templates. Apart from all these authentication formats, now-a-days biometric based applications came in to the technical world for providing the end users a more secured and sophisticated methodologies for authentication process across the secured systems. 

Every biometric system has it’s own advantage and disadvantages and in ideal case there is no perfect biometric identification system. The main usage and success of any biometric system depends on it’s unique features that can not be compared with few others and it’s user friendliness. The process being followed by any biometric system should be quick and easy to implement the features provided it, such as having a picture taken by a video camera, speaking into a microphone, or touching a fingerprint scanner. Below figure shows a typical biometric applications