Blind Navigation Obstacle Detection Dataset

Introduction

This dataset is designed for obstacle detection tasks in visually impaired navigation scenarios.

The dataset is annotated in YOLO format and can be used for real-time object detection research on mobile or embedded devices. It is mainly intended for lightweight YOLO-based models such as YOLOv8.

---

Dataset Features

• YOLO format annotations
• Designed for obstacle detection in blind navigation
• Lightweight and mobile-friendly
• Suitable for real-time inference tasks
• Supports Android deployment experiments

---

Classes

ID	Class Name
0	Bicycle
1	Bus
2	Car
3	Dog
4	Electric pole
5	Motorcycle
6	Person
7	Traffic signs
8	Tree
9	Uncovered manhole

---

Dataset Structure

dataset/
├── images/
│   ├── train/
│   ├── val/
│   └── test/
├── labels/
│   ├── train/
│   ├── val/
│   └── test/
└── data.yaml

---

Annotation Format

The dataset uses YOLO annotation format:

class_id x_center y_center width height

All coordinates are normalized to the image size.

---

Usage

Example training command using YOLOv8:

yolo detect train data=data.yaml model=yolov8n.pt epochs=100 imgsz=640

---

Applications

This dataset can be used for:

• Blind navigation assistance
• Mobile object detection
• Embedded AI systems
• Real-time obstacle avoidance
• Computer vision education and research

---

License

This dataset is released for academic and non-commercial research purposes only.

Commercial use is prohibited.

Redistribution without permission is prohibited.

---

Access Request

This dataset uses gated access.

Users must request access before downloading the dataset.

Please briefly describe your research purpose when submitting an access request.

---

Citation

If you use this dataset in your research, please cite:

@dataset{blind_navigation_dataset,
  title={Blind Navigation Obstacle Detection Dataset},
  author={ALTHISM},
  year={2026}
}