| Name | Number of images | Vehicle type | Image resolution | Annotations | Availability | Link |
|---|---|---|---|---|---|---|
| UFPR-ALPR | 4,500 | car, motorcycle | 1920x1080 | - vehicle position and type, - LP position and text - position of characters in LP |
By request | Link |
| RODOSOL-ALPR | 20,000 | car, motorcycle, bus, truck | 1280x720 | - vehicle type - LP layout - text - LP position |
By request | Link |
| AOLP | 2,049 | car | N/A | Not mentioned | By request | Link |
| CCPD (Chinese City Parking Dataset, ECCV) | 300k | car | 720x1160 | - LP area and tilt - LP coordinates - LP number - brightness and bluriness of LP |
Public | Link |
| CalTech Cars | 126 | car | 896 x 592 | not provided but a lot of people use it for testing | Public | Link |
| PKU | 4,000 | car | N/A | -LP segmenattions mask | Public | Link |
| Title | Date | Architecture | Train dataset | Eval datasets | Code | Link | Comments |
|---|---|---|---|---|---|---|---|
| An Efficient and Layout-Independent Automatic License Plate Recognition System Based on the YOLO detector | March 2021 | YOLO-based | Caltech Cars, EnglishLP, UCSD-Stills, ChineseLP, AOLP, SSIG-SegPlate, UFPR-ALPR | As for train | Configuration of the YOLO provided here | Link | none |
| Rethinking and Designing A High-performing Automatic License Plate Recognition Approach | June 2021 | CNN | CCPD, AOLP | CCPD, AOLP, PKUData, CLPD | Not found | Link | none |
| A Robust Real-Time Automatic License Plate Recognition Based on the YOLO Detector | April 2018 | YOLO | UFPR-ALPR, SSIG | UFPR-ALPR, SSIG | Official implementation not found | Link | none |
| License Plate Detection and Recognition in Unconstrained Scenarios | 2018 | CNN | subsets from CarsDataset, SSIG, AOLP manually annotated by the authors | OpenALPR, SSIG, AOLP | Code | Link | none |
| Persian License Plate Recognition System (PLPR) | 2023 | YOLO | IR-LPR, Iranis-Dataset, ILPR | Not mentioned, probably same as training | Code | No paper | none |
| LicensePlateDetector | Last commit 8 months ago | NeuralNetwork for recognizing characters, Connected Component Analysis | Personal dataset | Not mentioned | Code | No paper | none |
| Open-LPR | 2023 | - | - | - | Code | No paper | none |
| A Real-Time License Plate Detection Method Using a Deep Learning Approach | 2021 | YOLOv3 | - | - | Code (doesn't have OCR) | It has paper but need to enter with university account | none |
| A Flexible Approach for Automatic License Plate Recognition in Unconstrained Scenarios | 2021 | YOLO | -still figuring it out- | UFPR-ALPR, OpenALPR-BR, AOLP, CD-HARD, ?CCPD? | (it needs car cropped first. Other repos might need that too)(It also needs you to specify the type of vehicle: car, bus, truck)They used this: https://github.com/claudiojung/iwpod-net | Link | Localization works, but it doesn't include OCR. An ideea is to use Sergio's OCR from "License Plate Detection and Recognition in Unconstrained Scenarios" |
| Super-Resolution of License Plate Images Using Attention Modules and Sub-Pixel Convolution Layers | 2023 | super-resolution and sub-pixel convolutions | RODOSOL | Not mentioned | Code | Link | Just OCR - tested on RodoSol, also has git implementation. |