Naziva Septian, - (2024) PENGEMBANGAN SISTEM DETEKSI OBJEK PENUTUP WAJAH UNTUK PENGGUNA MESIN ATM DENGAN METODE YOLOV8. S1 thesis, Universitas Pendidikan Indonesia.
![]() |
Text
S_TEKOM_2008241_Title.pdf Download (2MB) |
![]() |
Text
S_TEKOM_2008241_Chapter1.pdf Download (1MB) |
![]() |
Text
S_TEKOM_2008241_Chapter2.pdf Restricted to Staf Perpustakaan Download (1MB) |
![]() |
Text
S_TEKOM_2008241_Chapter3.pdf Download (1MB) |
![]() |
Text
S_TEKOM_2008241_Chapter4.pdf Restricted to Staf Perpustakaan Download (3MB) |
![]() |
Text
S_TEKOM_2008241_Chapter5.pdf Download (1MB) |
![]() |
Text
S_TEKOM_2008241_Appendix.pdf Restricted to Staf Perpustakaan Download (4MB) |
Abstract
Penelitian ini dilatarbelakangi oleh kebutuhan peningkatan keamanan di ruang ATM melalui deteksi objek terlarang seperti helm, topi, dan kacamata hitam. Penelitian ini bertujuan untuk mengembangkan sistem deteksi objek untuk pengguna mesin ATM menggunakan metode You Only Look Once (YOLO), dengan fokus pada deteksi helm, kacamata hitam, dan topi yang sering digunakan untuk menyembunyikan identitas pelaku kejahatan. Alasan dilaksanakannya penelitian ini adalah karena meningkatnya kejahatan di ruang ATM, seperti skimming dan peretasan data nasabah, yang menunjukkan kerugian besar bagi perbankan Indonesia. Meskipun ada himbauan untuk tidak menggunakan helm, topi, dan kacamata hitam di ruang ATM, banyak pengguna yang masih mengabaikannya. Teknologi computer vision yang digunakan diharapkan dapat memberikan peringatan langsung kepada pengguna, meningkatkan keamanan, dan memastikan jarak pandang yang jelas terhadap individu yang sedang bertransaksi di ATM. Metode pengembangan sistem deteksi objek yang digunakan adalah AI Project Cycle yang meliputi problem scoping hingga deployment. Metode pengujian pengujian sistem deteksi objek dilakukan pada ruangan dan jarak yang berbeda. Pengembangan sistem deteksi objek pada Single Board Computer Jetson Nano menunjukan hasil evaluasi dengan akurasi deteksi sebesar 96.2%, dengan recall, precision, dan F1-score masing-masing 90.5%. Sistem ini mampu mendeteksi penggunaan helm, topi, dan kacamata hitam serta memberikan peringatan langsung kepada pengguna yang menggunakan objek tersebut. Namun, penelitian ini juga memiliki beberapa kekurangan, seperti penurunan akurasi pada jarak lebih jauh, kesalahan deteksi pada objek serupa, dan ketergantungan pada kondisi pencahayaan. ----------- This research is motivated by the need to increase security in the ATM room through the detection of prohibited objects such as helmets, hats, and sunglasses. This research aims to develop an object detection system for ATM machine users using the You Only Look Once (YOLO) method, focusing on the detection of helmets, sunglasses, and hats that are often used to hide the identity of criminals. The reason for conducting this research is due to the increasing crimes in ATM rooms, such as skimming and hacking of customer data, which shows great losses for Indonesian banks. Despite appeals not to use helmets, hats, and sunglasses in ATM rooms, many users still ignore them. The computer vision technology used is expected to provide immediate warnings to users, improve security, and ensure clear visibility of individuals who are transacting at the ATM. The object detection system development method used is AI Project Cycle which includes problem scoping to deployment. The test method of object detection system testing is carried out in different rooms and distances. The development of the object detection system on the Jetson Nano Single Board Computer showed evaluation results with detection accuracy of 96.2%, with recall, precision, and F1-score of 90.5% each. The system is able to detect the use of helmets, hats, and sunglasses and provide immediate warnings to users using these objects. However, this research also has some drawbacks, such as decreased accuracy at longer distances, detection errors on similar objects, and dependence on lighting conditions.
Item Type: | Thesis (S1) |
---|---|
Uncontrolled Keywords: | YOLO, deteksi objek, keamanan ATM, Jetson Nano, siklus proyek AI. YOLO, object detection, ATM security, AI project cycle. |
Subjects: | L Education > L Education (General) Q Science > QA Mathematics > QA76 Computer software T Technology > T Technology (General) |
Divisions: | UPI Kampus cibiru > S1 Teknik Komputer |
Depositing User: | Naziva |
Date Deposited: | 23 Aug 2024 03:18 |
Last Modified: | 29 Aug 2024 07:25 |
URI: | http://repository.upi.edu/id/eprint/120066 |
Actions (login required)
![]() |
View Item |