Traffic Detectors Using Image Processing Method (Pengesan Lalu Lintas Dengan Kaedah Pemprosesan Imej)
Most traffic detectors in service today work based on inductive loops buried under the road surface. Recently cameras and image processing with highspeed computers were beginning to be adopted for traffic surveillance system. This study was conducted to develop efficient traffic detector using image...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Penerbit UKM
2001
|
| Subjects: | |
| Online Access: | http://irep.iium.edu.my/37366/ http://irep.iium.edu.my/37366/ http://irep.iium.edu.my/37366/1/Jurnal_Kejuruteraan_2001.pdf |
| Summary: | Most traffic detectors in service today work based on inductive loops buried under the road surface. Recently cameras and image processing with highspeed computers were beginning to be adopted for traffic surveillance system. This study was conducted to develop efficient traffic detector using image-processing with a personal computer Traffic images were captured using ordinary analogue video camera and digitised using a video blaster. Four algorithms were developed i.e. algorithm for traffic counting, algorithm for vehicle classification, algorithm to determine the speed of the counted vehicles and algorithm for incident detection. The four algorithms were found to be sufficiently efficient to be executed by a personal computer to capture multiple road lane traffic data simultaneously with high accuracy both during daytime and during night.
(Kebanyakan pengesan lalu lintas yang ada sekarang menggunakan gelung induksi yang ditanam di bawah permukaan jalan raya. Semenjak akhir-akhir ini penggunaan kamera dan pemerosesan imej oleh komputer khas berkuasa tinggi telah mula digunakan untuk memantau lalu lintas. Kajian ini adalah untuk membangunkan pengesan lalu lintas yang cekap dengan kaedah pemperosesan imej oleh komputer peribadi. Imej lalu lintas diperolehi dengan menggunakan kamera analog dan didigitkan dengan menggunakan ‘video blaster'. Empat algoritnta telah dibangunkan, iaitu algoritma untuk mengira isi padu kenderaan, algoritma kenderaan, algoritma mendapatkan kelajuan kenderaan dan mengesan kejadian lalu lintas. Keempat-empat algoritma didapati cepat sehingga sebuah komputer peribadi berkemampuan untuk mendapatkatn data dari banyak lorong di jalan raya secara berasingan dalam masa serentak dengan ketepatan yang tinggi pada waktu siang atau malam). |
|---|