A First Look at Dataset Bias in License Plate Recognition

Laroca, Rayson; Santos, Marcelo; Estevam, Valter; Luz, Eduardo; Menotti, David

doi:10.1109/sibgrapi55357.2022.9991768

Cited by 10 publications

(6 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Creating a suitable dataset that covers all aspects of computer vision work is a major challenge. Bias in datasets, especially in ALPR datasets, has become a popular and controversial topic [2]. ALPR is a remarkable technology that depends on LPD as its first step.…”

Section: Background Datasetsmentioning

confidence: 99%

Iranian License Plate Recognition Using a Reliable Deep Learning Approach

Hatami,

Jamali,

Sadedel

2024

Scientia Iranica

View full text Add to dashboard Cite

The issue of Automatic License Plate Recognition (ALPR) has been a challenging one in recent years because of weather conditions, camera angle, lighting and different license plate characters. Due to advances in deep neural networks, it is now possible to recognize Iranian license plates using specific neural networks. The proposed method recognizes license plates in two steps. First, license plates are detected through the YOLOv4-tiny model, which is based on Convolutional Neural Network (CNN). Secondly, Convolutional Recurrent Neural Network (CRNN) and Connectionist Temporal Classification (CTC) are applied to recognize the license plate characters. For labels, one string of numbers and letters is enough without segmenting and labeling each separately. The proposed method boasts an average response time of 0.0074 seconds per image and 141 frames per second (fps) in the Darknet framework and 0.128 seconds per image in the TensorFlow framework for the License Plate Detection (LPD) part. This method has been proven to provide a highly accurate model with minimal storage space requirements, using less than 2MB for the Character Recognition (CR) model. There was an average accuracy of 75.14% and a response time of 0.435 seconds for the end-to-end process. The released code is available through GitHub.

show abstract

Section: Background Datasetsmentioning

confidence: 99%

Iranian License Plate Recognition Using a Reliable Deep Learning Approach

Hatami,

Jamali,

Sadedel

2024

Scientia Iranica

View full text Add to dashboard Cite

show abstract

“…4). Following [48], [49], we use 60% of the images for training/validation, while the remaining 40% are used for testing. Laroca et al [50] recently revealed that the PKU dataset (as well as several other datasets but not RodoSol-ALPR) has multiple images of the same vehicle/LP.…”

Section: A Setupmentioning

confidence: 99%

“…Following[12],[16],[48], we use the term "Brazilian" to refer to the layout used in Brazil prior to the adoption of the Mercosur layout.…”

mentioning

confidence: 99%

Super-resolution of license plate images using attention modules and sub-pixel convolution layers

Nascimento

Laroca

Lambert

et al. 2023

Computers & Graphics

View full text Add to dashboard Cite

“…4. We followed the approach of [41], [42], splitting 60% of the images for training and validation, and the remaining 40% for testing. To avoid bias, we grouped near-duplicates (distinct images of the same license plate) together, as recommended by Laroca et al [43].…”

Section: A Setupmentioning

confidence: 99%

“…This study focuses on the application of single-image super-resolution in the context of license plate recognition, as images from real-world surveillance systems are often characterized by low resolution and poor quality [7]- [9]. Although such challenging conditions are common in forensic applications, recent studies in license plate recognition have mainly concentrated on scenarios where the license plates are perfectly legible [10]- [14].…”

Section: Introductionmentioning

confidence: 99%

Super-Resolution Towards License Plate Recognition

Nascimento,

Laroca,

Menotti

2023

Anais Estendidos Da XXXVI Conference on Graphics, Patterns and Images (SIBRAPI Estendido 2023)

View full text Add to dashboard Cite

Recent years have seen significant developments in license plate recognition through the integration of deep learning techniques and the increasing availability of training data. Nevertheless, reconstructing license plates from low-resolution surveillance footage remains challenging. To address this issue, we propose an attention-based super-resolution approach that incorporates sub-pixel convolution layers and an Optical Character Recognition (OCR)-based loss function. We trained the proposed architecture on synthetic images created by applying heavy Gaussian noise followed by bicubic downsampling to high-resolution license plate images. Our results show that the proposed approach for reconstructing these low-resolution images substantially outperforms existing methods in both quantitative and qualitative measures. Our source code is publicly available at https://github.com/valfride/lpr-rsr-ext/.

show abstract

A First Look at Dataset Bias in License Plate Recognition

Cited by 10 publications

References 37 publications

Iranian License Plate Recognition Using a Reliable Deep Learning Approach

Iranian License Plate Recognition Using a Reliable Deep Learning Approach

Super-resolution of license plate images using attention modules and sub-pixel convolution layers

Super-Resolution Towards License Plate Recognition

Contact Info

Product

Resources

About