Overview

• Introduction
• Object Classes
  • Human
  • Wind-Sup-Board

Introduction

This dataset is designed to aid in the detection of humans and wind-sup-boards in aquatic environments. The goal is to accurately identify and annotate these objects within images captured from an aerial perspective. The following classes are included:

• Human: Individuals present in the water.
• Wind-Sup-Board: Stand-up paddleboards, with or without sails.

Object Classes

Human

Description

The human class includes all visible parts of a person in the water. Typically, these appear as small figures often partially submerged or floating, with limbs or heads protruding above/below the water.

Instructions

• Annotate any visible part of the human body, including heads, arms, and legs.
• If a person is only partially visible due to water coverage, include visible part and infer the rest for the bounding box.
• Ensure the bounding box encompasses all visible body parts without extending into the water unnecessarily.
• Disambiguate from floating debris or equipment by focusing on the shape and form of human limbs.
• Humans can be on wind-sup-boards, still label them.

Wind-Sup-Board

Description

Wind-sup-boards are elongated, often oval-shaped boards used for stand-up paddling, sometimes equipped with a sail. They appear as larger floating objects in the water.

Instructions

• Annotate the full extent of the board, from tip to tail, including any visible sail.
• Include the paddle if it is laying on the board.
• Boards almost always have human's on them, so ensure to create separate human bounding boxes too.

                        @misc{
                            human-detection-in-floods-a6aun-5xvpd-2hvjd-sbyy_dataset,
                            title = { human-detection-in-floods-a6aun-5xvpd-2hvjd-sbyy Dataset },
                            type = { Open Source Dataset },
                            author = { Roboflow 100-VL },
                            howpublished = { \url{ https://universe.roboflow.com/rf100-vl/human-detection-in-floods-a6aun-5xvpd-2hvjd-sbyy } },
                            url = { https://universe.roboflow.com/rf100-vl/human-detection-in-floods-a6aun-5xvpd-2hvjd-sbyy },
                            journal = { Roboflow Universe },
                            publisher = { Roboflow },
                            year = { 2025 },
                            month = { mar },
                            note = { visited on 2025-03-23 },
                            }