Inmates Tracking Based on Face and Visual Markers
| dc.contributor.author | DRICI, ADEM | |
| dc.contributor.author | HAMIDOUCHE, SALAH | |
| dc.date.accessioned | 2024-09-17T10:35:51Z | |
| dc.date.available | 2024-09-17T10:35:51Z | |
| dc.date.issued | 2024-06 | |
| dc.description.abstract | Traditional prison identification methods, such as RFID, wristbands, and fingerprinting, have many defects and shortcomings. For this reason, we propose a new system that uses facial recognition and ArUco marker technologies, adapted to prisoner uniforms. Each prisoner is given a unique ArUco marker ID that enables accurate identification even if the face is obstructed. Our system will help track prisoners throughout the prison by placing cameras at strategic locations, entrances, corridors, and jail squares, ensuring constant surveillance and optimizing detection capability. This idea gives better security with improved efficiency and reliability and can completely revolutionize prison surveillance | en_US |
| dc.identifier.issn | MM/812 | |
| dc.identifier.uri | http://10.10.1.6:4000/handle/123456789/5377 | |
| dc.language.iso | en | en_US |
| dc.publisher | UNIVERSITY BBA | en_US |
| dc.subject | RFID,Tracking Visual Markers | en_US |
| dc.title | Inmates Tracking Based on Face and Visual Markers | en_US |
| dc.type | Thesis | en_US |
Files
Original bundle
1 - 1 of 1
- Name:
- memoire_Final Salah_hm1.pdf
- Size:
- 3.58 MB
- Format:
- Adobe Portable Document Format
- Description:
- Ensuring the safety and security within prisons relies heavily on effective inmate tracking systems. Our exploration begins with both traditional and modern methods used to monitor inmates. Traditional methods include video surveillance, visual and direct inspections, and frequent communication between officers. These approaches ensure continuous monitoring and quick responses to any issues. In addition to these methods, biometric tools such as facial recognition, fingerprint identification, and retina and iris scans have become integral to modern inmate tracking. Behavioral characteristic tools, like signature and voice recognition, further enhance monitoring capabilities. The use of RFID technology allows real-time location tracking of inmates, significantly improving monitoring accuracy and efficiency. Facial recognition technology is a sophisticated system that involves several key steps: face detection, face analysis, image-to-data conversion, and comparison and identification. This chapter delves into each of these steps, outlining their roles and significance in the overall process. In our system, we employ the Haar Cascade Algorithm for face detection due to its efficiency and reliability. Following detection, we utilize the Local Binary Pattern (LBP) method for face recognition. This chapter provides a comprehensive explanation of these techniques, highlighting their advantages and how they contribute to the accuracy and effectiveness of our facial recognition system. In our system, we utilize Aruco Markers to track inmates within correctional facilities. Here, we explore the application of visual markers in both traditional and modern contexts, emphasizing 63 the use of Aruco marker technology to enhance inmate monitoring. In prisons, visual markers such as Identification Numbers, Barcodes or QR Codes, and Colored Bracelets are used for identification and management. Identification numbers are assigned to each inmate, while barcodes or QR codes facilitate quick scanning and digital record-keeping. Colored bracelets help in identifying different categories of inmates, such as those with medical conditions or varying security levels. Modern applications of visual markers include QR Codes, which are widely used for accessing information and verifying identities, and Aruco Markers, which we use in our system as a backup to track inmates in cases where their faces do not appear. Identification of Aruco markers begins by explaining their structure and decoding process. This involves capturing an image, applying adaptive thresholding, extracting squares, validating these squares through inner codification, and finally extracting and displaying the marker's ID. Integrating facial recognition technology with ArUco markers enhances prisoner tracking and monitoring within correctional facilities. We begin with a class diagram illustrating the structure of our program, providing a clear overview of its components and their interactions. This integrated approach leverages two advanced technologies: facial recognition and ArUco markers. ArUco markers, affixed to the prisoners' suits, serve as unique identifiers for each inmate. The program is designed to track inmates using both facial recognition and ArUco marker detection. When a prisoner's face is not detected, the ArUco marker acts as a backup identifier, ensuring continuous tracking. Additionally, when an ArUco marker is detected, the program displays the prisoner's name alongside the marker's ID. A critical feature of the program is its ability to detect and alert authorities if a prisoner exchanges suits with another inmate. If the face recognition result does not match the ArUco marker ID, an alert is triggered, indicating a mismatch between the face's name and the marker's name. This mechanism helps maintain security and enforces disciplinary action against offenders who attempt to deceive the system. 64 We provided practical examples demonstrating the program in action, accompanied by images and performance metrics, showcasing the accuracy and reliability of the tracking system. Additionally, we covered the tools and technologies used in the development of the program, such as Python, OpenCV, and pandas. We also presented a use case diagram that explains the method and workflow of the integrated system, offering a visual representation of how the program operates. This comprehensive understanding of the integrated facial recognition and ArUco marker system highlights its innovative approach to enhancing prisoner tracking and security within correctional facilities.
License bundle
1 - 1 of 1
No Thumbnail Available
- Name:
- license.txt
- Size:
- 1.71 KB
- Format:
- Item-specific license agreed to upon submission
- Description: