PoTATO
收藏资源简介:
PoTATO数据集由佐治亚理工学院创建,包含12,380个标记的塑料瓶图像,具有丰富的偏振信息。数据集大小为2448×2048像素,通过安装在Kingfisher USV上的Triton 5.0MP TRI050S1-QC相机在Lake Symphonie上采集。数据集创建过程中,通过手动操作船只,确保了背景、光照和太阳位置的多样性。该数据集主要用于水面上漂浮物体的检测,旨在解决户外光照条件和水面反射对物体检测精度的影响问题。
The PoTATO dataset was created by the Georgia Institute of Technology. It consists of 12,380 labeled plastic bottle images with rich polarization information. All images in the dataset have a resolution of 2448 × 2048 pixels. The dataset was collected at Lake Symphonie using a Triton 5.0MP TRI050S1-QC camera mounted on a Kingfisher USV. During the dataset creation process, manual vessel operation was adopted to ensure diversity in backgrounds, lighting conditions, and solar positions. This dataset is primarily used for floating object detection on water surfaces, aiming to address the adverse impact of outdoor lighting conditions and water surface reflections on object detection accuracy.

- 1PoTATO: A Dataset for Analyzing Polarimetric Traces of Afloat Trash Objects佐治亚理工学院 · 2024年



