Synthetic image generation for training deep learning-based automated license plate recognition systems on the Brazilian Mercosur standard

Silvano, Gilles; Ribeiro, Vinícius Campos Tinoco; Greati, Vitor Rodrigues; Bezerra, Aguinaldo; Silva, Ivanovitch; Endo, Patrícia Takako; Lynn, Theo

doi:10.1007/s10617-020-09241-7

Cited by 12 publications

(6 citation statements)

References 49 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Assuming that the image is rotated in a clockwise direction, the offset of the horizontal axis X axis direction, relative to the horizontal direction, and the vertical axis can be calculated using the cos and sin functions, respectively; the pixels in the vertical axis direction are calculated quantitatively with sin and cos functions, respectively, relative to the horizontal axis and vertical axis offsets of the original point [13,14]. erefore, whether it is a forward or reverse rotation, all the pixel values of the image can be calculated by reasoning using the above process, so as to determine the position of the pixel, but it should be noted that except for the initial pixel needing floating point multiplication, the rest of the subsequent calculations are completed by floating point addition, which saves calculation time and improves calculation efficiency.…”

Section: Principle Of Image Derotationmentioning

confidence: 99%

Digital Image Automatic Recognition and Tracking Based on Electronic Derotation Algorithm

Gao

2021

Advances in Multimedia

View full text Add to dashboard Cite

With the continuous development of my country’s social economy, the ways to acquire images have become more and more abundant. How to effectively process, manage, and mine images has become a major and difficult problem in research. In view of the difficult problem of image recognition, the electronic derotation algorithm is introduced in this study, by combing and monitoring the edge features, establishing a corresponding sample database, analyzing the edge features of the image, and performing effective and stable tracking, so as to realize the automatic recognition and tracking of the digital image. The simulation experiment results show that the electronic derotation algorithm is effective and can support the automatic recognition and tracking of digital images.

show abstract

Section: Principle Of Image Derotationmentioning

confidence: 99%

Digital Image Automatic Recognition and Tracking Based on Electronic Derotation Algorithm

Gao

2021

Advances in Multimedia

View full text Add to dashboard Cite

show abstract

“…These images are fictitious, obtained at the computational level and created in order to reproduce the real scenario. By simulating the data in a virtual environment, it is possible to influence parameters that have an impact on the images: light scenarios, camera positions, environments and actions [ 57 ]. Synthetic datasets can therefore be used for computer vision tasks starting from image classification over instance segmentation or anomaly detection.…”

Section: Methodsmentioning

confidence: 99%

Sea Mine Detection Framework Using YOLO, SSD and EfficientDet Deep Learning Models

Munteanu

Diana

Zamfir

et al. 2022

Sensors

View full text Add to dashboard Cite

In the context of new geopolitical tensions due to the current armed conflicts, safety in terms of navigation has been threatened due to the large number of sea mines placed, in particular, within the sea conflict areas. Additionally, since a large number of mines have recently been reported to have drifted into the territories of the Black Sea countries such as Romania, Bulgaria Georgia and Turkey, which have intense commercial and tourism activities in their coastal areas, the safety of those economic activities is threatened by possible accidents that may occur due to the above-mentioned situation. The use of deep learning in a military operation is widespread, especially for combating drones and other killer robots. Therefore, the present research addresses the detection of floating and underwater sea mines using images recorded from cameras (taken from drones, submarines, ships and boats). Due to the low number of sea mine images, the current research used both an augmentation technique and synthetic image generation (by overlapping images with different types of mines over water backgrounds), and two datasets were built (for floating mines and for underwater mines). Three deep learning models, respectively, YOLOv5, SSD and EfficientDet (YOLOv5 and SSD for floating mines and YOLOv5 and EfficientDet for underwater mines), were trained and compared. In the context of using three algorithm models, YOLO, SSD and EfficientDet, the new generated system revealed high accuracy in object recognition, namely the detection of floating and anchored mines. Moreover, tests carried out on portable computing equipment, such as Raspberry Pi, illustrated the possibility of including such an application for real-time scenarios, with the time of 2 s per frame being improved if devices use high-performance cameras.

show abstract

“…In this sense, Silvano et al (2021) presented a methodology that couples synthetic images of Mercosur LPs with real-world images containing vehicles with other LP layouts. A model trained exclusively with synthetic images achieved promising results on 1,000 real images from various sources; however, it is difficult to assess these results accurately since the test images were not made available to the research community.…”

Section: Related Workmentioning

confidence: 99%

On the Cross-dataset Generalization in License Plate Recognition

Laroca,

Cardoso,

Lucio

et al. 2022

Preprint

View full text Add to dashboard Cite

Automatic License Plate Recognition (ALPR) systems have shown remarkable performance on license plates (LPs) from multiple regions due to advances in deep learning and the increasing availability of datasets. The evaluation of deep ALPR systems is usually done within each dataset; therefore, it is questionable if such results are a reliable indicator of generalization ability. In this paper, we propose a traditional-split versus leave-one-dataset-out experimental setup to empirically assess the cross-dataset generalization of 12 Optical Character Recognition (OCR) models applied to LP recognition on nine publicly available datasets with a great variety in several aspects (e.g., acquisition settings, image resolution, and LP layouts). We also introduce a public dataset for end-to-end ALPR that is the first to contain images of vehicles with Mercosur LPs and the one with the highest number of motorcycle images. The experimental results shed light on the limitations of the traditional-split protocol for evaluating approaches in the ALPR context, as there are significant drops in performance for most datasets when training and testing the models in a leave-one-dataset-out fashion.

show abstract

Synthetic image generation for training deep learning-based automated license plate recognition systems on the Brazilian Mercosur standard

Cited by 12 publications

References 49 publications

Digital Image Automatic Recognition and Tracking Based on Electronic Derotation Algorithm

Digital Image Automatic Recognition and Tracking Based on Electronic Derotation Algorithm

Sea Mine Detection Framework Using YOLO, SSD and EfficientDet Deep Learning Models

On the Cross-dataset Generalization in License Plate Recognition

Contact Info

Product

Resources

About