APPENDIX 1 – L7 COMPUTER VISION PROBLEM OUTLINE FORM
Maintaining the attendance plays a very crucial role in all the organisations. Each organisation has their own method. The traditional method of taking attendance either manually like old paper/ file-based approaches and few organisations in the recent days have adopted new biometric methods of managing the attendance. The traditional attendance systems and the recent bio-metric techniques are non-reliable and time consuming. Time taken to mark attendance in 60 minutes of class is usually 15 minutes and the traditional attendance management systems are not fool proof. Besides that, some students might just give the attendance and the leaves the class before it finishes. We are trying to build an automated system which manages the attendance system by Face Detection and Face Recognition in a live environment and marks their corresponding attendance in the system. Face Detection detects the difference between an object and a face and captures only faces data. Face Recognition comes into place after a face is detected and this detected face is passed onto the face recogniser to confirm the identity of the face detected. This whole process is carried out on the live streaming of video captured while entering the class. This solution will help us in improving the efficiency of attendance management system by minimising the duplicate entries, false information, the time taken to mark the attendance and gets rid of human error.
The aim of this project is to create a system that automates the strategy of analysing faces computationally and mark attendance of the students present within the class into the database.
The following measurable objectives have been identified:
Attendance is a state of being present at a place or at an event. Attendance of every student is being maintained by all the educational institutions these days. The time spent for the manual attendance of each class is approximately 45 minutes for all the lectures in a day. Here our main objectives are to reduce the time taken for marking the attendance and eliminate the duplicate entries in the database.
A live streaming camera is used to capture the video of a classroom, the model built on the analysed data is used and applied to detect and recognise the face and mark their corresponding attendance. This whole process is carried out in a live environment, on pre-recorded videos and on images of classroom too. The precious time taken to mark the attendance manually can be saved as there is no need of manual attendance system in this case and the system will take the attendance of the students automatically without the need for human intervention and store it in the database.
Furthermore, we are trying the eliminate the duplicate entries in an attendance. It is possible for another person to fake an attendance and make duplicate entries. We are overcoming it by automated attendance management systems. These systems will be completely fool proof and reliable as they minimise the human errors. It should also be able to recognise time stamp the in time and exit time of a student in every lecture.
In order to deliver this project, the following steps are planned:
Data Preparation: The input image data is captured in multiple angles of every student face and stored with their corresponding student details containing Name, Identity number, Date of Birth, Sex, Email address. These are labelled and stored as raw data in the database for further analysis
Face Detection: Paul Viola and Michael Jones proposed the method to detect face. The Voila-Jones method is used for its high accuracy, low false detection, real time (for practical applications which is at least 2 frames per second must be processed).
AdaBoost and Artificial Neural Networks. We use this to train node classifiers on a Haar-like feature set to improve the generalization ability of the node classifier. The face detection performance of the face detector is improved consequently.
Face Recognition: Principal component analysis using eigenfaces, elastic bunch graph matching using the Fisherface algorithm, linear discriminant analysis, the hidden Markov model, the neuronal motivated dynamic link matching and the multilinear subspace learning using tensor representation.
We will also be checking out the other different algorithms and techniques used for facial detection and recognition and improve their performance.
The attendance capturing process begins with capturing the image or video of every person whose attendance needs to be recorded. Face recognition uses the unique features of a human’s face in order to recognise and identify that person. These features are captured by a camera in the class at a safe distance. The process starts with face detection, where faces from a video or an image are detected and passed to the recognition stage where the face are identified by comparison with an existing database of faces.
The attendance of the recognised student faces are recorded and stored in a database as CSV files for the attendance check. After facial detection and recognition, student attendance algorithms are combined to form the attendance management system. Implement techniques to get rid of duplicate entries.
DescriptionIn this final assignment, the students will demonstrate their ability to apply two ma
Path finding involves finding a path from A to B. Typically we want the path to have certain properties,such as being the shortest or to avoid going t
Develop a program to emulate a purchase transaction at a retail store. Thisprogram will have two classes, a LineItem class and a Transaction class. Th
1 Project 1 Introduction - the SeaPort Project series For this set of projects for the course, we wish to simulate some of the aspects of a number of
1 Project 2 Introduction - the SeaPort Project series For this set of projects for the course, we wish to simulate some of the aspects of a number of