YOLOv8 Bike Analog Odometer Detection Model

Model Description

This custom YOLOv8 model has been trained to detect and read analog odometers specifically on bikes.

Training Dataset

The dataset used to train this model consists of images of bike analog odometers collected from various sources. These images were annotated with bounding boxes around the odometer regions. The dataset includes odometers from different bike models under various lighting conditions, angles, and image qualities to ensure robustness.

Dataset Format

Images: JPEG/PNG formats
Annotations: YOLO format (text files with bounding box coordinates)
Dataset:

Training Details

Model: YOLOv8
Framework: Ultralytics YOLOv8
Epochs: 100

Augmentation Techniques

Random rotation
Horizontal/vertical flips
Random brightness and contrast adjustments

Usage

This model can be used to automatically detect analog odometers in bike images. Here’s how to use it:

Example

from ultralytics import YOLO

# Load the trained model
model = YOLO('weights/last.pt')

# Perform detection on an image
results = model('path/to/your/image.jpg')

# Display results
results.show()

Applications

Odometer Reading

Limitations

The model may not perform well in extreme lighting conditions (e.g., very low light or strong glares).
Limited to analog odometers; does not detect digital odometers.
And also I used 640x640 size 202 images so in some cases the model not work well

Future Work

Extend the model to handle digital odometers.
Improve performance under challenging environmental conditions.
Collect more diverse data for training to enhance robustness.

License

Apache 2.0

Citation

If you use this model in your research or projects, please cite it as follows:

@misc{arun-arunisto2024yolov8_bike_odometer,
  title={YOLOv8 Bike Analog Odometer Detection Model},
  author={Arun Arunisto},
  year={2024},
  howpublished={\url{https://huggingface.co/Arunisto/odometer-analog-detector}},
}