Automatic Attendance System Using Matlab
Uploaded by
IJRASETPublicationsAutomatic Attendance System Using Matlab
Uploaded by
IJRASETPublicationsInternational Journal for Research in Applied Science & Engineering Technology (IJRASET)
ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 6.887
Volume 6 Issue V, May 2018- Available at [Link]
Automatic Attendance System using Matlab
Neha1, Nikunj Saxena2, Anuradha Bhasin3
1
Student, ECE, BPIT, GGSIPU
2
Student, ECE, BPIT, GGSIPU
3
Associate professor, BPIT, GGSIPU
Abstract: In this paper we are proposing an automatic attendance system which can be used in every organization to mark the
attendance of employees. The main application of Automatic attendance system is seen in teaching institutions, where the
attendance of students has to be regularly monitored on daily basis. The method developed provides a secure and effective may
recording attendance. Automatic attendance system uses mainly two algorithms i.e. Viola Jones Algorithm and Local binary
pattern algorithm. Viola Jones algorithm is used for face detection and Local binary pattern is used for the feature extraction
and face recognition.
Keywords: Viola Jones algorithm, Local binary pattern, Haar like features, Integral images, Adaptive boost, Cascading, Image
segmentation, Thresholding.
I. INTRODUCTION
Face recognition is the analysis of characteristics of a person's face image taken from a camera. It measures overall facial feature
like distances between eyes, length of nose, distance between mouth, and length between jaw edges. These measurements are stored
in a database and used for comparison when a user stands in front of the camera. Camera can be placed a distance of 2 feet to ten
feet from the person to be identified, without waiting for long periods of time or do anything more than look at the camera. Face
recognition has become a popular area of research and successful application of image processing.
Traditional method of attendance is manual to maintain the record of lectures, salary and work hours. This attendance is generally
maintained using attendance sheet. The traditional marking methods are monotonous & time consuming. Manually recorded
attendance can be easily manipulated. Because these problems being faced in manual attendance system, some of the organizations
have shifted to biometric technique but biometric system serves one person at a time.
In this paper we have proposed an automatic attendance system that can tackle all the problems faced in manual attendance system
as well as in biometric system. The proposed automatic attendance system consists of a high resolution digital camera to monitor the
workplace. It is embedded on a micro-controller based motor system so that it can rotate in left & right directions. The data or
images obtained by the camera are used by computer programmed system for further calculation. The images are compared with
reference images of each of the employees or students to mark the corresponding attendance.
The challenges of face recognition lie in the inherent variability arising from face characteristics like illumination, quality like
resolution geometry like distance and viewpoint and signal to noise ratio, and image content like background, occlusion and
disguise [2]. It is the ability to establish a subject’s identity based on his facial characteristics. Automatic face recognition has been
extensively studied over the past two decades due to its number of application domains, such as access control, visual surveillance
II. RELATED WORK
In automatic attendance system we have studied and identified face images. For this we have used two main algorithms Voila Jones
algorithm for face detection and Linear binary pattern for feature extraction and face recognition. A brief discussion of the
algorithms used is given:
A. Viola Jones algorithm
The Viola-Jones face detection algorithm is the first framework based on object detection that provides good detection rates in real-
time is given by Paul Viola & Michael Jones in the year of 2001. This algorithm has been implemented in a software ‘Matlab’ [7].
This algorithm basically consist of below mentioned algorithms.
1) Haar like feature: Haar-like feature is defined as the difference of the sum of pixels of areas inside the rectangle, which can be
at any position and scale within the original image. This difference is then used to categorize subsections of an image. This
modified feature set is called 2-rectangle feature. Viola and Jones also defined and used 3-rectangle features and 4-rectangle
features. [1]
©IJRASET: All Rights are Reserved 1058
International Journal for Research in Applied Science & Engineering Technology (IJRASET)
ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 6.887
Volume 6 Issue V, May 2018- Available at [Link]
2) Integral Images : Integral images is used to obtain the sum of the pixels in a rectangular subset of a grid quickly and efficiently.
Integral images calculate all the pixel above and to the left of (x, y) and assign sum to the (x, y). [1]
3) Adaptive Boost : Adaboost or adaptive boost is a predictive algorithm for classification and regression. Adaboost is mainly used
for over fitting only. Adaboost is called adaptive because it uses multiple iterations to generate a single portion of the image. [1]
4) Cascading: The basic principle of cascading is to scan the whole image. It may be possible that it can contain more than one
face so cascading scans and detect face in the picture. [1]
B. Local Binary Pattern
Local binary pattern is used for the feature extraction and face recognition. In this algorithm an image is divided into multiple parts
and on each part LBP is applied. First we make 3 X 3 top left corner window is then each pixel is converted into equivalent grey
level. After this we compare centre pixel value is with all the neighbourhood values and as result if centre pixel value is greater than
surrounding pixel then it is 1 otherwise [Link] next step , the 8-bit code is converted into decimal equivalent and further this procedure
is repeated for the whole image. After this all the decimal values are plotted on histogram. [4]
Image segmentation is mainly used in partitioning a digital image into multiple segments (sets of pixels, also known as super
pixels).The goal of segmentation is to modify the representation of an image into more meaningful and easier to analyse. Image
segmentation is typically used to locate objects and boundaries (lines, curves, etc.) in images. More precisely, image segmentation is
the process of assigning a label to every pixel in an image such that pixels with the same label share certain characteristics.
Thresholding is the simplest method of image segmentation. From a grayscale image, Thresholding can be used to create binary
images. The simplest thresholding methods replace each pixel in an image with a black pixel if the image intensity is less than some
fixed constant or a white pixel if the image intensity is greater than that constant. [4]
III. SYSTEM OVERVIEW
A. Enrollment Phase
1) Image Acquisition: For Image acquisition, camera is used to capture an image of the employees working in an organization.
2) Face Detection and Face cropping: For face detection cascading algorithm is used. Further face cropping is done using imcrop
(image) command that separates the facial area from the rest of the background image. This faces is stored in the database.
3) Feature Extraction: Feature extraction is done by using linear binary pattern algorithm. Feature extraction is helpful for face
detection and recognition.
4) Database: Extracted features of employees or persons will be stored in Database.
©IJRASET: All Rights are Reserved 1059
International Journal for Research in Applied Science & Engineering Technology (IJRASET)
ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 6.887
Volume 6 Issue V, May 2018- Available at [Link]
B. Routine Attendance After Enrollment Phase
Routine attendance will start after extracting the features of all the images stored in the database. We have followed the same steps
as given in enrollment phase i.e. image acquition, Face detection, and face cropping
1) Face recognition: Face recognition is done by using the Local binary pattern by taking 3x3 block a time and comparing central
block with surrounding blocks and compare with features stored in the database. If after face recognition face of the person is
matched in the database then update the attendance otherwise show the dialog box with invalid person.
IV. RESULT
A. Test image for enrollment phase
Twenty images were tested from which only two of them are included in this paper. These two images indicates successful
enrollment of the persons.
Figure 1 Test image 1
Figure 2 Test image 2
B. Test Image for Routine Attendance phase
Following are the test results of the routine attendance phase. These two images indicates that their attendance has been marked
successfully.
Figure 3 Test image 1
©IJRASET: All Rights are Reserved 1060
International Journal for Research in Applied Science & Engineering Technology (IJRASET)
ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 6.887
Volume 6 Issue V, May 2018- Available at [Link]
Figure 4 Test image 2
V. CONCLUSION
The complete system is implemented in MATLAB. We have implemented the algorithm on MATLAB 2013a version. This
attendance system shows the use of facial recognition techniques. The project was developed in reference to the various aspects of
its need & applications in the present scenario. Implementation of our project is developed as in accordance with our proposed block
diagram.
REFERENCE
[1] Khan Suhel, Zakariya Hussain, Khan Shoeb, Pathan Nazim, Mubashir Khan, Apeksha Gopale, “Automated Attendance System Using Face Recognition”, Vol.
4, Issue 4, April 2016.
[2] Jomon Joseph, K. P. Zacharia, “Automatic Attendance Management System Using Face Recognition”, International Journal of Science and Research (IJSR).
[3] Abhishek Jha, “Class Room Attendance System Using Facial Recognition System”, The International Journal of Mathematics, Science, Technology and
Management (ISSN: 2319-8125) Vol. 2 Issue3
[4] Mrunmayee Shirodkar, Varun Sinha, Urvi Jain, Bhushan Nemade, “Automated Attendance Management System using Face Recognition”, International Journal
of Computer Applications (0975 – 8887) International Conference and Workshop on Emerging Trends in Technology (ICWET 2015
[5] Jageshvar K. Keche, Vikas K. Yeotikar, Manish T. Wanjari, Dr. Mahendra P. Dhore, “Human Face Recognition Based on PCA Method using MATLAB”,
National Conference on Advanced Technologies in Computing and Networking ATCON-2015 Special Issue of International Journal of Electronics,
Communication & Soft Computing Science and Engineering, ISSN: 2277-9477.
[6] Dr. Nita Thakare1, Meghna Shrivastava2, Nidhi Kumari3, Neha Kumari4, Darleen Kaur5, Rinku Singh6, “Face detection and recognition for automatic
attendance system”, Dr. Nita Thakare, International Journal of Computer Science and Mobile Computing, Vol.5 Issue.4, April- 2016, pg. 74-78.
[7] Vikram, K., and S. Padmavathi. "Facial parts detection using Viola Jones algorithm." Advanced Computing and Communication Systems (ICACCS), 2017 4th
International Conference on. IEEE, 2017.
©IJRASET: All Rights are Reserved 1061
Common questions
Powered by AIVariability in facial recognition systems can arise from external factors such as illumination changes, facial occlusions, variations in head poses, and background noise, which can distort the captured image, leading to recognition errors . Additionally, internal system limitations, such as insufficiently detailed facial databases and inadequate algorithmic robustness, can contribute to inaccuracies. Mitigation strategies include enhancing image capture conditions by stabilizing lighting, using advanced preprocessing techniques to normalize input images, improving algorithm robustness through deep learning models, and expanding the databases with diverse facial images to increase system flexibility and accuracy under varying conditions .
The automatic attendance system utilizes two primary algorithms: Viola-Jones and Local Binary Pattern (LBP). The Viola-Jones algorithm is used for face detection, capitalizing on its ability to provide good detection rates in real-time. It employs techniques like Haar-like features, integral images, adaptive boost, and cascading to efficiently locate faces within images . The Local Binary Pattern (LBP) algorithm is used for feature extraction and face recognition. It works by dividing an image into multiple parts, comparing pixel values within each part, and converting these comparisons into a numerical format for easy classification . Both algorithms together enable the system to accurately and efficiently capture and recognize faces for attendance marking.
Face recognition systems face challenges due to inherent variability caused by factors such as illumination, resolution quality, geometry (distance and viewpoint), signal-to-noise ratio, and image content like background, occlusion, and disguise . Variability in these factors can lead to differences in how a person's face is perceived by the system, potentially causing difficulties in matching the captured face with those stored in the database. Such discrepancies can lead to errors in recognition, thereby affecting the overall performance and reliability of the attendance system .
The automatic attendance system improves upon traditional manual attendance methods by offering a more secure, efficient, and less labor-intensive alternative. Traditional methods are often time-consuming and prone to errors and manipulation, as they typically involve manual data entry on attendance sheets. In contrast, the automatic system uses facial recognition to mark attendance, significantly reducing the time and effort required for attendance-taking while minimizing human errors and fraudulent entries . Additionally, the use of facial recognition allows for continuous monitoring and attendance recording without the need for individuals to manually report their presence .
Integral images are used in the Viola-Jones algorithm to enable rapid and efficient calculation of Haar-like features, which are critical for face detection. By employing integral images, the algorithm can quickly compute the sum of pixel values within specific rectangular areas of the image, allowing it to assess multi-scale features without extensive computational effort . This optimization greatly accelerates the face detection process, making it feasible to operate in real-time scenarios, as the computational time for evaluating any feature can be reduced to a few operations irrespective of the feature's size or location .
The cascading technique in the Viola-Jones algorithm is a robust method designed to quickly and accurately detect faces within images by processing different stages sequentially. This technique operates by scanning the entire image and applying multiple stages of feature detection in increasing order of complexity. Each stage consists of a collection of weak classifiers that work together to form a strong classifier. If a section of the image passes through all stages, it is classified as containing a face; otherwise, it is discarded early in the process, reducing processing time and computational load . The cascading technique is significant as it optimizes the detection process, ensuring that only the promising regions are subjected to intensive processing, thereby enhancing real-time detection capability .
Image segmentation is crucial in breaking down a digital image into multiple segments to enhance its analysis and interpretation. In the context of automatic attendance systems, segmentation is employed to isolate and identify important facial features by partitioning the image into meaningful areas . This process aids in locating objects and boundaries, such as face outlines and facial features, which are necessary for accurate facial recognition. By labeling each pixel and ensuring that pixels in identical segments share certain characteristics, image segmentation enables the system to accurately match facial features against those stored in the database, improving both recognition accuracy and speed .
Biometric systems primarily serve one person at a time and can be inefficient and cumbersome in processing large groups, leading to delays in attendance recording . Additionally, they can be subject to failures due to changes in physical characteristics, such as injuries or wear, raising false negatives. The proposed automatic attendance system addresses these limitations by utilizing face recognition technology, which can simultaneously capture and process multiple subjects without physical contact through passive observation . This not only speeds up the process of attendance marking but also reduces downtime and potential inaccuracies caused by direct interaction with the biometric devices.
The enrollment phase of the automatic attendance system is crucial for ensuring accurate face recognition in routine use. During this phase, image acquisition is performed by capturing images of employees, and face detection is carried out using the cascading algorithm. The detected faces are then cropped from the background images and the features are extracted using the Local Binary Pattern algorithm, which enables detailed and distinctive recognition of individual facial features . These extracted features are stored in a database, serving as a reference against which future images are compared. This ensures that during routine attendance, the system can accurately recognize and validate individuals by matching their current facial features against the stored data .
The camera system plays a central role in the implementation of the automatic attendance system by capturing real-time images of individuals within its operational field. Positioned typically between two to ten feet from the subjects, the camera uses high resolution to ensure clarity in the images captured, which is essential for accurate face detection and recognition . These images serve as the input for further processing by the computer system where facial features are analyzed and compared with the database to record attendance. The camera's ability to rotate left and right, if mounted on a motor, can enhance the coverage area, thus supporting the dynamic environment of workplaces or classrooms .