Boat detection in satellite imagery MASATI-V2

The MASATI-V2 Dataset is a curated collection of labeled satellite images designed for boat detection and localization in oceanic and coastal environments. This dataset is widely used in training and evaluating machine learning models, particularly YOLO-based object detection frameworks.

Key Features of the MASATI-V2 Dataset

• Total Images: 4,522 labeled satellite images of the ocean and coastlines.
• Annotations: Bounding boxes highlighting the precise location of boats.
• Pre-Split Data: The dataset is pre-split into training, testing, and validation sets, optimized for YOLO models.
• Image Source: Satellite imagery obtained from Microsoft® Bing™ Maps.
• Supported Models: Compatible with YOLOv10 Nano and other object detection architectures.

Applications of MASATI-V2

The dataset is tailored for tasks such as:

1. Maritime Surveillance: Detect and monitor boats for security and research purposes.
2. Environmental Monitoring: Analyze coastal activities and their environmental impact.
3. AI and Machine Learning: Train robust object detection models for satellite imagery.

Licensing and Usage

• License: For non-profit research and educational purposes only.
• Terms of Use: Data sourced from Microsoft® Bing™ Maps. Refer to their terms of use for details.
• Citation Requirement: If using the dataset, please reference the original research publications as specified.