FACE RECOGNITION SYSTEM

ABSTRACT:-

            In present times, face popularity has come to be one of the great technology for laptop vision. Face popularity is continually a very hard venture in laptop vision, illumination, pose, facial expression. Face popularity tracks goal objects in stay video photographs enthusiastic about a video digital digicam. In simple words, it’s far a machine software for robotically identifying someone from a nevertheless picture or video frame. In this paper we proposed an automatic face popularity machine. This software primarily based totally on face detection, characteristic extraction and popularity algorithms, which robotically detects the human face whilst the individual in the front of the digital digicam spotting him. We used KLT Algorithm, Viola-Jones Algorithm face detection which come across human face the usage of Haar cascade classifier, but digital digicam is constantly detecting the face each frame, PCA set of rules for characteristic selection. We follow a version combining to in shape the geometric traits of the human face.

EXISTING SYSTEM :-

Face detection The trouble of face popularity is all approximately face detection. Face detection is similarly categorized as face detection in pictures and real-time face detection. In this venture we will try and come across faces in nevertheless pictures via way of means of the use of photo invariants. To try this it’d be beneficial to look at the greyscale depth distribution of a mean human face. The following ‘common human face became made out of a pattern of human faces. A definitely scaled colour map has been used to spotlight grayscale depth differences. The gray-scale differences, which can be invariant throughout all the pattern faces are strikingly apparent. The eye-eyebrow area appear to usually comprise darkish depth (low) graylevels at the same time as nostril brow and cheeks comprise vibrant depth (high) gray levels. After a splendid deal of experimentation, the researcher observed that the following regions of the human face had been appropriate for a face detection machine primarily based totally on photo invariants and a deformable template. The above facial area plays nicely as a foundation for a face template, likely due to the clean divisions of the vibrant depth invariant area via way of means of the darkish depth invariant regions. Once this pixel area is positioned via way of means of the face detection machine, any precise area required may be segmented primarily based totally at the proportions of the common human face. Face popularity Human face popularity may be divided into strategies: geometrical capabilities and template matching. a) Face popularity the use of geometrical capabilities It entails computation of a hard and fast of geometrical capabilities including nostril width and length, mouth function and chin shape, etc. from the image of the face we need to understand. This set of capabilities is then matched with the capabilities of acknowledged individuals. A appropriate metric including Euclidean distance (locating the nearest vector) may be used to discover the nearest match. The benefit of the use of geometrical capabilities as a foundation for face popularity is that popularity is viable even at very low resolutions and with noisy pictures (pictures with many disorderly pixel intensities). Although the face can’t be regarded in detail its overall geometrical configuration may be extracted for face popularity. The method’s predominant drawback is that computerized extraction of the facial geometrical capabilities is very hard.

Geometrical capabilities (white) which may be used for face popularity b) Face popularity the use of template matching This is comparable the template matching method used in face detection, besides right here we aren’t looking to classify an photo as a ‘face’ or ‘non-face’ however are looking to understand a face. The foundation of the template matching approach is to extract complete facial regions (matrix of pixels) and compare those with the saved pictures of acknowledged individuals. Once once more Euclidean distance may be used to discover the nearest match. Algorithm for face Recognition Adding the photo to the database 1. Get the photo. 2. Get the FaceDetector item. 3. Apply the FaceDetector item to the photo to extract the capabilities of detectedface. 4. Add the photo to the database. Comparing the enter photo with the database of pictures 1. Get the photo. 2. Get the FaceDetector item. 3. Apply the FaceDetector item to photo and extract the capabilities. 4. Compare the photo with the database.

PROPOSED SYSTEM:-

• Face Tracking

The objective of this set of rules is to discover item of face in actual time and to maintain monitoring of the identical item. Here we use the schooling samples photos of different items of your preference to be detect and music through schooling classifier. Face monitoring is part of face reputation machine. Here we can use a few machine algorithms to select out out specific, unique information about a human’s face.

• Face Detection

 In [1] This face detection procedure definitely verifies the photo is face photo or now no longer. Detection procedure definitely works on Haar Cascade classifier. Object Detection using Haar characteristic- based classifiers is an powerful item detection technique proposed Paul Viola and Michael Jones. It is gadget studying primarily based totally technique in which a cascade characteristic is educated from photos. It is used to discover items in different photos.

• Haar Cascade Classifier Features

 In [2] Here we calculated, the primary characteristic decided on seems to attention at the belongings that the location of the eyes in often darker than the location of the nostril and cheeks. The second characteristic selected is primarily based totally on the attention is darker characteristics than the bridge of the nostril. However, you do now no longer want the identical window that applies for your cheeks and different locations. face reputation machine that does taking pictures the photo of face characteristic detection, extraction, storing and matching. But the issue takes place to put the transmission traces in the locations in which the topography is bad. The authors proposed a machine primarily based totally on actual-time face reputation that is reliable, stable and fast, and calls for development in extraordinary lighting fixtures conditions.

MODULES:-

• Registration: Students who register in a portal for the first time submit their personal data, their ID card and their photo, which is stored in the database and verified using their photo before the exam.
  
  
• Face recognition: A webcam is installed in the a student’s computer or  front camera, when the student takes a test on a face recognition recognizes the student  and if the face matches the stored facial image, the student is verified and allowed to take the exam.During the exam, the student’s image is continuously captured and if the face does not match the stored image, their record is saved in the database. Multiple face detection: If there is more than one person  in the picture, this is also recorded in the database.
  
  
• Head Position Detection: For MCQ-based exams that do not require pencil and paper, students’ head position is analyzed and if it appears that a student is looking at the other side of the screen, your dataset is also analyzed be saved.
  
  
• Cell Phone Detection – If a student is found with a cell phone, this will also be recorded as bad practice in the database.Misconduct and will be logged.

ADVANTAGES :-

• Payments and Authentication. Face Recognition can be used in order to facilitate payments and make them more convenient for both, the business and the customer. …
• Access Control and Security.
• Attendance Monitoring.
• Airport services.
• Age control.
• Accuracy Issues.
• Privacy concerns.
• Data Misusage.

HARDWARE AND SOFTWARE REQUIREMENTS:-

HARDWARE:-

• Processor: Intel Core i3 or more.
• RAM: 4GB or more.
• Hard disk: 250 GB or more.

SOFTWARE:-

• Operating System : Windows 10, 7, 8.
• Python
• Anaconda
• Spyder, Jupyter notebook, Flask.
• MYSQL
• Ganache