Parking Lot Location Information Recognition Dataset

Description： This dataset focuses on visual location indicators found within parking facilities, including area codes, floor numbers, and parking spot identifiers. The images (1920×1080, RGB) were collected from indoor and outdoor parking environments using a variety of devices such as vehicle-mounted cameras, handheld devices, and surveillance systems (exact sources to be confirmed).

The dataset supports applications such as automatic parking guidance, vehicle localization assistance, vision-based indoor navigation, and OCR-based perception systems for underground structures. It enables machine learning models to recognize and extract structured information from real-world parking lot scenes.

Each image is annotated with text region bounding boxes and transcription labels, following OCR-oriented annotation formats in JSON or other structured formats (to be confirmed). The labeled entities include various combinations such as:

Pure numbers: e.g., "123", "07"

Chinese + numbers: e.g., "B区05号"

Letters + numbers: e.g., "A5", "L2-3"

Composite expressions: e.g., "P3 地下二层 B12"

Captured signs are located on walls, columns, ramps, and ground directly in front of parking spots, covering both natural and artificial lighting (including nighttime scenes). The visual styles vary across fonts, colors, materials, and layouts, reflecting the diversity of real-world environments in ground-level, underground, and multi-level parking lots.

This dataset offers a reliable foundation for developing robust vision-based parking assistance and guidance systems. It facilitates OCR tasks involving both structured and semi-structured location data in complex indoor environments.

This dataset contains over 20,000 high-quality annotated images. The estimated price is approximately €8,000, and can be adjusted based on delivery format, licensing scope, and customization requirements.

Keywords: parking floorcode areacode location OCR signage Sector: Smart Parking, Intelligent Transportation Systems, Indoor Navigation, OCR Vision Metrics: OCR Accuracy, Position Detection Precision, Character Recognition Rate Entities: Area Numbers, Floor Numbers, Parking Spot Codes Geographies: N/A Price Range (USD): 4000–20000 Source type: Synthetic data generation from simulation environment, human annotation

Sample Images: