Freiburg Forest

弗莱堡森林数据集是使用 Viona 自主移动机器人平台收集的，该平台配备了用于捕获多光谱和多模态图像的相机。该数据集可用于评估用于分割、检测、分类等的不同感知算法。所有场景均以 20 Hz 的频率记录，相机分辨率为 1024x768 像素。由于阴影和太阳角度在获取图像的质量中起着至关重要的作用，因此在三个不同的日子里收集了数据，以便在照明条件上有足够的可变性。机器人每天行走约 4.7 公里。数据集创建者为 6 个类别提供手动注释的逐像素地面实况分割掩码：障碍物、小径、天空、草、植被和 Void。

The Freiburg Forest Dataset was collected using the Viona autonomous mobile robot platform, which is equipped with cameras for capturing multispectral and multimodal images. This dataset can be used to evaluate various perception algorithms for tasks such as segmentation, detection, classification, and more. All scenarios were recorded at a frequency of 20 Hz, with the camera resolution set to 1024×768 pixels. Since shadows and solar angles play a crucial role in the quality of acquired images, data was collected on three different days to ensure sufficient variability in lighting conditions. The robot traveled approximately 4.7 kilometers each day. The dataset creators provide manually annotated pixel-wise ground-truth segmentation masks for six categories: obstacles, trails, sky, grass, vegetation, and Void.

搜集汇总

背景与挑战

背景概述

Freiburg Forest数据集是由弗莱堡大学于2016年发布的森林场景图像数据集，专为评估感知算法如语义分割、检测和分类而设计。它通过Viona自主移动机器人平台采集，包含多光谱和多模态图像，以20Hz频率和1024x768像素分辨率记录，覆盖不同光照条件，并提供6个类别的逐像素手动注释分割掩码，适用于计算机视觉研究。