Realistic License Plate Restoration and Recognition Dataset (RLPR)

Description

[Dataset Description] This dataset was constructed to facilitate the development of methods aimed at improving image quality and recognition accuracy of license plates captured from dashcam videos. It contains paired low-quality license plate video sequences and corresponding high-quality target images, enabling quality enhancement model training and evaluation. Each sample includes a sequence of 31 low-quality license plate frames alongside a single high-quality license plate image that serves as the target (pseudo-ground truth) for quality enhancement. This dataset can be used for training models to enhance image quality and includes recognition labels for assessing recognition accuracy. Additionally, the dataset provides outputs obtained using the MF-LPR² model proposed in our paper, offering a benchmark for performance comparison. [Dataset Contents] 1) Low-Quality (LQ) License Plate Image Sequences - 31 frames per sequence. 2) High-Quality (HQ) License Plate Images (Pseudo-Ground Truth) - Includes both the original images and versions cropped to the Region of Interest (ROI). - ROI coordinates provided. 3) License Plate Recognition Label - For privacy reasons, only the numerical characters of the license plate are provided. 4) MF-LPR² Model Super-Resolution (SR) Results - Outputs generated by the proposed MF-LPR² model 5) Homography Transformation Coordinates - Coordinates for aligning low- and high-quality images - Utility tool is also provided [Citation] When using this dataset, please cite the following paper: Na, K., Oh, J., Cho, Y., Kim, B., Cho, S., Choi, J., & Kim, I. (2025). MF-LPR2: Multi-frame license plate image restoration and recognition using optical flow. Computer Vision and Image Understanding, 256, 104361. [Disclaimer] All Korean Hangul characters in the license plate regions have been pseudonymized using targeted blurring techniques. In the low-resolution (LR) image sequences, the frames have been cropped to the license plate region to minimize the exposure of the surrounding vehicle or environment. In some cases, small portions of the vehicle may remain visible due to cropping margins. The recognition labels contain only numerical values. No GPS metadata, timestamps, vehicle make/model information, or environmental background data is included in this dataset. This dataset is released strictly for non-commercial, academic research purposes. By accessing or using this dataset, you agree to the following terms: * You will not attempt any form of re-identification or misuse of the data. * You will not use the dataset for surveillance, law enforcement, or any commercial applications. * You will cite the original paper when publishing any results that use this dataset. * You will comply with all applicable data protection and privacy laws. The authors and their affiliated institutions assume no legal responsibility for any use of this dataset that extends beyond its intended academic research purpose.

Steps to reproduce

1. Collection of Low-Quality Sequences: We extracted 31-frame sequences from 1,052 dash cam videos, selecting frames with visible license plates that were sufficiently low in quality for restoration. Only sequences containing low-quality license plates that require enhancement were included. 2. Generation of High-Quality Reference Images (Pseudo GT): From each sequence, we selected the frame with the highest quality license plate as a "pseudo ground truth (GT)" image. This serves as the target for quality enhancement of the low-quality sequence. 3. License Plate Detection and Cropping: We used a DeepLabV3 model, based on a ResNet-101 backbone and fine-tuned on 130 manually annotated road images, to detect the license plate regions in each frame. We then manually refined the detected regions and expanded the bounding boxes to ensure no important details were missed during cropping. 4. Alignment and Transformation of Pseudo GT Images: To align the high-quality reference image with the low-quality frames, we applied a homography transformation by marking the four corner points of the license plates in both the pseudo GT and the low-quality frames. This allowed us to accurately align the high-quality image with the target frames, focusing only on the license plate regions during evaluation.

Institutions

Categories

Funding

Related Links

Licence