Simulated Disaster Victim (SDV1 & SDV2) dataset

Simulated Disaster Victim dataset consists of images and video frames containing simulated human victims in cluttered scenes along with pixel-level annotated skin maps. The simulation was carried out in a controlled environment with due consideration towards the health of all the volunteers. To generate a real effect of a disaster, Fuller’s earth is used which is skin-friendly and does not cause harm to humans. It created an effect of disaster dust over the victims in different situations. The victims included one female and four male volunteers. The objective of this dataset generation was to capture images having humans lying in different postures in a catastrophic scenario. The clutter is brought in by using construction material, such as bricks, cement bags, dust, Fuller's earth, stones, etc. The datasets have a variety of variations including color, scale, pose, illumination, motion blur, and occlusion. The dataset consists of two parts SDV1 and SDV2. SDV1 consists of 128 images and ground truth maps. SDV2 video dataset has 15 video clips. The SDV2 dataset has pixel-level annotations of skin maps around which skin objectness windows are formed for random frames for each video. To the best of our knowledge, there is no dataset publicly available testing algorithms used for assisting rescuers in locating victims. This dataset will serve as a benchmark dataset benefiting future research efforts in this direction.

模拟灾难受害者数据集包含在杂乱场景中模拟的人类受害者图像和视频帧，以及像素级标注的皮肤图。模拟在严格控制的环境中完成，充分考虑到所有志愿者的健康。为生成灾难的真实效果，使用了对人体友好的 Fuller's earth，它不会对人类造成伤害，并在受害者身上创造了不同情境下的灾难尘埃效果。参与者包括一名女性和四名男性志愿者。生成此数据集的目标是捕捉到人类在灾难场景中以不同姿势躺卧的图像。通过使用建筑材料，如砖块、水泥袋、灰尘、Fuller's earth、石块等，引入了杂乱元素。该数据集具有多种变化，包括色彩、尺度、姿态、照明、运动模糊和遮挡等。数据集分为两部分：SDV1和SDV2。SDV1包含128张图像和地面实况图。SDV2视频数据集包含15个视频片段。SDV2数据集对每个视频的随机帧周围的皮肤图进行了像素级标注，并形成了皮肤目标窗口。据我们所知，目前尚无公开的数据集用于测试辅助救援人员定位受害者的算法。本数据集将作为基准数据集，为未来在此方向上的研究工作提供益处。