OverView: Using yolo object detection and homography to localize people overboard for automonous USV's search and rescue missions with integration in moos-ivp middleware
This repository contains a custom Maritime Search and Rescue (SNR) dataset, specifically curated for object detection tasks using the YOLOv5 model. Additionally, the repository includes a MOOS-ivp wrapper for YOLOv5 and for OpenCV homography to facilitate seamless integration with the MOOS-ivp framework.
The custom Maritime SNR dataset comprises annotated images and corresponding object labels necessary for training and evaluating an object detection model. The dataset focuses on maritime scenarios and contains a variety of objects commonly encountered during SNR operations, such as lifebuoys, life jackets, boats, and more.
Please refer to the dataset/ directory for more information on the dataset structure and instructions on how to use it for training your YOLOv5 model.
YOLOv5 is a popular real-time object detection model known for its efficiency and accuracy. In this repository, we provide a pre-trained YOLOv5 model that has been fine-tuned using the custom Maritime SNR dataset. You can use this model to perform object detection on new maritime SNR images or fine-tune it further on your specific dataset.
Instructions on how to use the YOLOv5 model for inference and training can be found in the yolov5/ directory.
The MOOS-ivp wrapper for YOLOv5 allows you to seamlessly integrate the YOLOv5 object detection capabilities into your MOOS-ivp applications or systems. By using this wrapper, you can leverage the power of YOLOv5 for real-time object detection in maritime SNR scenarios within the MOOS-ivp framework.
Details on how to use the MOOS-ivp wrapper for YOLOv5 can be found in the moos_ivp_wrapper/ directory.
We welcome contributions from the community to enhance the dataset, improve the YOLOv5 model, or extend the MOOS-ivp wrapper for YOLOv5. Please refer to the CONTRIBUTING.md file to learn about the contribution guidelines.
The code and dataset in this repository are provided under the MIT License.
Note: Below is the code version of the README. Please use it for your repository's README.md file.
# Custom Maritime Search and Rescue Dataset
This repository contains a custom Maritime Search and Rescue (SNR) dataset, specifically curated for object detection tasks using the YOLOv5 model. Additionally, the repository includes a MOOS-ivp wrapper for YOLOv5 to facilitate seamless integration with the MOOS-ivp framework.
## Dataset
The custom Maritime SNR dataset comprises annotated images and corresponding object labels necessary for training and evaluating an object detection model. The dataset focuses on maritime scenarios and contains a variety of objects commonly encountered during SNR operations, such as lifebuoys, life jackets, boats, and more.
Please refer to the `dataset/` directory for more information on the dataset structure and instructions on how to use it for training your YOLOv5 model.
## YOLOv5 Model for Object Detection
YOLOv5 is a popular real-time object detection model known for its efficiency and accuracy. In this repository, we provide a pre-trained YOLOv5 model that has been fine-tuned using the custom Maritime SNR dataset. You can use this model to perform object detection on new maritime SNR images or fine-tune it further on your specific dataset.
Instructions on how to use the YOLOv5 model for inference and training can be