PDD: A Pixel-Level Road Pavement Deformation Dataset for Semantic Segmentation of Road Surface Distresses

The Pavement Deformation Dataset (PDD) was developed to support research on automatic pavement deformation analysis, pavement condition assessment, and intelligent road inspection in real-world environments. The dataset is designed for semantic segmentation and includes five pavement deformation categories: Depression, Rutting, Upheaval, Pothole, and Patching. Polygon-based annotations enable accurate delineation of irregular pavement deformation boundaries and support quantitative pavement analysis. RGB pavement images were collected from asphalt roads across Lampung Province, Indonesia, including Bandar Lampung City, South Lampung, East Lampung, Central Lampung, Metro City, and Pesawaran Regency. Data acquisition was conducted using vehicle-mounted Logitech C925e Webcam, Intel RealSense D455 RGB-D camera, and GoPro HERO10, together with a Samsung Galaxy A55 smartphone for handheld image acquisition. Images were automatically captured at 2-second intervals while the survey vehicle traveled at approximately 10–40 km/h, whereas additional still images and representative video frames were acquired using the GoPro HERO10 and Samsung Galaxy A55. Data collection was performed under diverse road, weather, and illumination conditions using an oblique-view camera configuration (30°–45°) with approximately 4–8 m pavement coverage. A total of 6,332 original RGB pavement images were collected and manually screened to ensure image quality and annotation suitability. Only images containing visible pavement deformation were retained. Since a single image may contain multiple pavement deformation instances, polygon annotations were manually created using CVAT according to predefined class definitions. The retained images were standardized to a uniform resolution of 1024 × 1024 pixels while preserving the original RGB information. No image enhancement or data augmentation was applied, and all images included in the dataset represent original field observations. The dataset is organized according to the standard YOLO segmentation directory structure and comprises 4,450 training (70%), 1,282 validation (20%), and 600 test (10%) images. The repository contains RGB pavement images, polygon annotation files in YOLO segmentation format, and a data.yaml configuration file defining the dataset paths and class names. The dataset provides a standardized resource for semantic segmentation benchmarking and supports pavement condition assessment, automated road inspection, infrastructure monitoring, and intelligent transportation systems. It can also be reused or reannotated for reproducible computer vision research and related applications.