AULIA, NISARIFAH (2026) PERANCANGAN SISTEM IDENTIFIKASI OTOMATIS NOMOR MESIN DAN RANGKA MENGGUNAKAN KAMERA ENDOSCOPE DAN TEKNOLOGI OCR. Diploma thesis, POLITEKNIK KESELAMATAN TRANSPORTASI JALAN.
|
Text (ABSTRAK)
23031016-KKW-ABSTRAK.pdf Download (869kB) | Preview |
|
|
Text (BAB 1)
23031016-KKW-BAB_1.pdf Download (137kB) | Preview |
|
![[img]](http://eprints.pktj.ac.id/style/images/fileicons/text.png) |
Text (BAB 2)
23031016-KKW-BAB_2.pdf Restricted to Registered users only Download (440kB) | Request a copy |
|
|
Text (BAB 3)
23031016-KKW-BAB_3.pdf Restricted to Registered users only Download (609kB) | Request a copy |
|
|
Text (BAB 4)
23031016-KKW-BAB_4.pdf Restricted to Registered users only Download (982kB) | Request a copy |
|
|
Text (BAB 5)
23031016-KKW-BAB_5.pdf Download (81kB) | Preview |
|
|
Text (LAMPIRAN)
23031016-KKW-BAB_LAMPIRAN.pdf Restricted to Registered users only Download (3MB) | Request a copy |
Abstract
The inspection of engine numbers and chassis numbers of motor vehicles at Regional Technical Implementation Units for Vehicle Testing (UPTD PKB) has conventionally been carried out using the rubbing method with paper and pencil. This method presents several limitations, including unclear results due to the concealed position of engine numbers in confined areas, potential injury risks for inspectors, and the absence of digitally integrated inspection records. This study aimed to design and evaluate the performance of an automatic identification system for engine numbers and chassis numbers of motor vehicles using an endoscope camera and Optical Character Recognition (OCR) technology. The research was conducted at the UPTD PKB of the Sragen Regency Transportation Agency, Central Java. The system was developed on the Android platform using Android Studio, integrating a 7 mm diameter, 720P resolution endoscope camera connected via USB OTG, image preprocessing using OpenCV (grayscale, CLAHE, bilateral filter, and adaptive thresholding), offline text recognition using Google ML Kit Text Recognition, and local data storage based on SQLite. The system was evaluated through three testing methods: OCR accuracy testing using the Character Error Rate (CER) formula, black box testing with 16 test scenarios, and a time-efficiency comparison between the conventional method and the system method on 50 vehicle samples. Testing results demonstrated that OCR accuracy for chassis numbers reached 98.24% and for engine numbers 90.43%, both exceeding the field tolerance threshold of 85–90%. All 16 black box scenarios were declared Valid. Time efficiency for chassis number identification averaged 49.42% and for engine numbers 40.27% faster compared to the conventional method, with a cumulative saving of over 24 minutes per 50-vehicle testing cycle. The system is concluded to have been successfully designed, fully functional, and viable for implementation as a safer, more accurate, and more efficient alternative for motor vehicle number identification. Keywords: automatic identification; endoscope camera; optical character recognition; Google ML Kit; motor vehicle inspection
| Item Type: | Thesis (Diploma) |
|---|---|
| Uncontrolled Keywords: | automatic identification; endoscope camera; optical character recognition; Google ML Kit; motor vehicle inspection |
| Subjects: | T Technology > T Technology (General) T Technology > TL Motor vehicles. Aeronautics. Astronautics |
| Divisions: | Teknologi Otomotif > Teknologi Otomotif |
| Depositing User: | 23031016 23031016 |
| Date Deposited: | 24 Jun 2026 01:09 |
| Last Modified: | 24 Jun 2026 01:09 |
| URI: | http://eprints.pktj.ac.id/id/eprint/4737 |
Actions (login required)
 |
View Item |
