Attendance System Essay Example

larger gatherings like lectures or conferences, it poses challenges. The disadvantages of the manual attendance system include time wasted on waiting for student responses and the use of paper, as well as the inconvenience of not generating attendance reports promptly.

The attendance report that is shared through NITR webmail is outdated by two months. In order to address this issue, it is crucial for us to adopt an automatic online attendance management system. Therefore, we are presenting an feasible framework for managing attendance. This framework is comprised of three main components: Hardware/Software Design, Attendance Management Approach, and On-line Report Generation.

The LCD display will show the rolls of students whose attendance is marked. The computer software will connect to the fingerprint scanner, LCD, and network. It will input and process the fingerprint, extracting features for matching. After matching, the attendance records of the students will be updated in the database.

The computer software starts the process by getting the teacher's fingerprint and determining the Subject ID and Current Semester using either the teacher's ID or manual input. If the teacher does not enter the classroom, attendance marking will not start. Also, if there is a delayed entrance of over 20 minutes, attendance will not be registered.

The location does not have the necessary range for wireless LAN, so data will be stored on the chip. Attendance results will only be updated when the portable device is near the nodes storing reports. However, we can still update all records online using mobile networks from different companies.

The current 3G network enables clients to update attendance records automatically without physically going to nodes. The

attendance data will be retrieved from a central database repository using the 3G mobile network. The responsibility of designing such a portable device falls on embedded system engineers. This device can have either a touchscreen input/display or buttons with an LCD display. A specially-designed software will run on the device. Before allowing students to mark attendance, teachers will first verify their fingerprints on the device.

After confirming the teacher's identity, the software will prompt for details about the course and other necessary information regarding the class being taught. The software will also inquire about the time limit for which attendance should not be marked, which can be adjusted based on the teacher's preference. However, we recommend a default value of 25 minutes to discourage late arrivals of students. This process will only take a few seconds. Subsequently, students will receive a device for fingerprint identification and attendance marking. Following this, the teacher can proceed with their lecture.

The device will be circulated among students without recorded attendance. After 25 minutes or the teacher's designated time limit, the device will no longer accept attendance inputs. When class ends, the teacher will retrieve the device and conclude the lecture. The primary purpose of this device's software is to identify students through fingerprint recognition, generate reports, and transmit them to servers via the 3G network. Additionally, it allows for downloading and updating of the database from a central repository.

There are various types of student attendance management systems, including the RFID based system and the GSM-GPRS based system. Each system has its own advantages and disadvantages. However, our unique system stands out as it not

only saves teaching time but is also portable. Unlike GSM-GPRS based systems that rely on the class's location to mark attendance, our system can adapt to changes in schedule or location, ensuring accurate recording of attendance.

The problem with RFID systems is the need for students to carry cards and install detectors. However, our system eliminates these issues by using fingerprints as recognition criteria, preventing proxies. Additionally, attendance marking can be done anywhere and anytime if portable devices are used. Thus, our student attendance system is the ideal choice for implementation at NITR.

We improve the quality of fingerprints by eliminating noise and extracting features such as minutiae. If the sets of minutiae match those in the database, the fingerprint is identified. After matching, post-matching actions may be taken, such as displaying candidate details and recording attendance.

By utilizing a minutiae extraction algorithm on the image's background regions, it is possible to extract noisy and false minutiae. This process involves discarding these background regions through segmentation, resulting in a more reliable extraction of minutiae points. To accomplish this, we will use a variance thresholding method. The background regions typically have minimal grey-scale variance, while the foreground regions exhibit significantly higher variance. Initially, the image is divided into blocks, and the grey-scale variance is calculated for each block within the image.

The Gabor filter has a beneficial feature of having zero DC component, resulting in a filtered image with zero mean pixel value. Hence, to binarize the image, a global threshold of zero can be applied. Binarization includes examining the grey-level value of each pixel in the enhanced image. If the value surpasses the

predetermined global threshold, the pixel value becomes one; otherwise, it is set to zero.

The result of binarisation is a binary image that consists of background valleys and foreground ridges. Thinning is a morphological operation used to eliminate certain foreground pixels from the binary images. A standard thinning algorithm with two subiterations is applied for this operation.

Using the 'thin' operation of the bwmorph function in MATLAB, the algorithm allows users to access it. It begins each subiteration by analyzing the neighborhood of each pixel in the binary image. Based on specific pixel-deletion criteria, the algorithm determines if a pixel should be removed or not. These subiterations continue until no more pixels can be removed.

The use of a key helps save time by avoiding the need to match minutiae sets that do not have a complete match. This is particularly useful in large databases where matching the entire minutiae set for each enrolled fingerprint would be unnecessary and wasteful. There are two types of keys - simple and complex - and this project specifically uses the simple key.

Because the reference point detection algorithm is not accurate or perfect, we utilized the centroid of all minutiae to create the key. The complex key is more intricate and contains more information. It stores a vector of minutiae where each subsequent minutiae is the closest to the previous one, beginning with either the reference point or the centroid of all minutiae.

The matching algorithm, which will be discussed in a later section, is used for both key matching and full matching. This algorithm matches the key of the

query fingerprint with the keys in the database. If a match is found, full matching occurs. If no match is found, the query key is matched with the next M keys and so on.

The process involves comparing the ith minutiae in the query set with k unmatched minutiae in the sample set. Both sets must be sorted according to their distance from the reference point or centroid. The ith minutia from the query set is then compared with the top k unmatched minutiae in the database minutiae set.