Deep convolutional neural network-based recognizer for tricycle franchise permit
No Thumbnail Available
Date
2024-01
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Don Mariano Marcos Memorial State University - South La Union Campus
Abstract
Vehicle recognition is a critical component of surveillance and control systems. This includes utilizing license plates with unique alphanumeric codes for identification. However, manually recognizing license plates for all parked or passing vehicles is a labor-intensive task. To address this, a real-time automated franchise permit recognition system is proposed in this study using deep convolutional neural networks (CNNs).
The proposed system includes franchise permit detection using YOLOv7 and recognition using a multilayer perceptron. The results demonstrate that the YOLOv7 model achieved an average precision of 92.70 percent in detecting franchise permits with an IOU threshold of 0.50, while the MLP model achieved 95.78 percent accuracy in recognizing franchise permits. The confusion matrix for digit classification also shows high accuracy rates ranging from 93 percent to 98 percent for digits 0-9. Overall, the proposed deep CNN-based recognizer for tricycle franchise permit showed promising results in terms of mean average precision, accuracy, and inference time.
Description
Abstract Only
Keywords
Citation
Imatong, A.T. (2024). Deep convolutional neural network-based recognizer for tricycle franchise permit. [Unpublished Master's Thesis]. Don Mariano Marcos Memorial State University – South La Union Campus, Agoo, La Union. Lakasa ti Sirib, DMMMSU Institutional Repository.
Collections
Total visits
| views | |
|---|---|
| Deep convolutional neural network-based recognizer for tricycle franchise permit | 0 |
File Visits
| views | |
|---|---|
| MT-SLUC-2024-CGS-MSCS-ImatongAT-ABS.pdf | 1 |