stevenalbert10/Toyota-Corolla-Car-Parts

---
license: apache-2.0
---
# ARabeitak Car Parts Object Detection Dataset

## Dataset Summary
The **ARabeitak Car Parts Object Detection Dataset** is a small, task-oriented image dataset of car engine compartments captured to train and evaluate object detection models.

Images were collected as part of the graduation project:
> **“ARabeitak: Markerless Augmented Reality Platform for User Support – ‘ARabeitak’ Car Maintenance App as a Case Study”**
> *Judy Wagdy, Laila Hany, Rana Raafat, Steven Albert*
> *Supervised by Dr. Essam Eliwa, Eng. Nada Ayman*
> *Submitted June 2023, in partial fulfillment of the requirements for the degree of BSc in Computer Science.*

The dataset focuses on common under-the-hood components that a non-expert driver may need to locate during basic car maintenance tasks. It was used to train object detection models powering the ARabeitak mobile application, which overlays AR guidance directly on the car’s engine bay.

---

## Dataset Description
### Use Cases
- Training and evaluating **object detection** models for:
- Interactive AR guidance for car maintenance.
- Intelligent car manuals / remote support.
- Educational applications for basic automotive literacy.

### Data Modality
- **Images:** RGB photographs of car engine bays and related parts.

### Structure
- Images are organized into **7 folders** by car part / view.
- **7 folders correspond to the labeled object classes** listed below.

---

## Classes
The dataset is categorized into the following **7 object classes**.
The numbers below indicate the current **count of object images** per class:

| Class Label | Count |
|----------------------|-------|
| Air Filter | 206 |
| Battery | 269 |
| Coolant Reservoir | 296 |
| Engine | 212 |
| Engine Cover | 187 |
| Reservoir Cap | 11 |
| Windshield Fluid | 171 |

---
## Data Instances
Each instance includes:
- **Image**
- Under-the-hood photo taken from a user’s point of view.
- Real-world conditions: dust, shadows, varying illumination, etc.

## Collection Process
### Source

- Photos captured from a Toyota Corolla 2015 engine bay.
- Shot with consumer-grade (Samsung Galaxy S20+) mobile device in natural conditions (no studio setup).