Malaysian vehicle license plate recognition using deep learning and computer vision
License plate recognition has become one of the popular topics under deep learning researches. There are many deep learning models and the suitable model for this project chose according to the ability to meet the system operation requirements such as speed, accuracy and precision of the outcome. Th...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Conference or Workshop Item |
| Language: | English English |
| Published: |
Springer Science and Business Media Deutschland GmbH
2022
|
| Subjects: | |
| Online Access: | http://umpir.ump.edu.my/id/eprint/39530/1/Malaysian%20Vehicle%20License%20Plate%20Recognition%20Using%20Deep%20Learning.pdf http://umpir.ump.edu.my/id/eprint/39530/2/Malaysian%20vehicle%20license%20plate%20recognition%20using%20deep%20learning%20and%20computer%20vision_ABS.pdf http://umpir.ump.edu.my/id/eprint/39530/ https://doi.org/10.1007/978-981-16-8690-0_88 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | License plate recognition has become one of the popular topics under deep learning researches. There are many deep learning models and the suitable model for this project chose according to the ability to meet the system operation requirements such as speed, accuracy and precision of the outcome. Therefore, YOLO (You Only Look Once) model was used which is fast in processing the more images and produce the output at a single look. YOLO is an algorithm designed for multi object detection in a single neural network where it only sees once and process to detect object as many as possible in a picture. In this paper, YOLOv3 is use to detect the position of car registration plate. Next, image warping and slicing applied to straighten the image so it will be easy to feed into character recognition process. Then, the PyTesseract will be used to read the characters from the image together with RegEx function to eliminate the weak predictions from the PyTesseract results. The results obtained from this approach achieved 100% accuracy in recognizing vehicle car plate from 5 video collected from Universiti Malaysia Pahang (UMP) main entrance security gate CCTV system. |
|---|