随着之后的使用,可能会有一些补充
indoor datasets NYU V2 [5], S3DIS [6], ScanNet [7] and outdoor datasets Semantic.3D [8], KITTI [9], vKITTI [10]
semantic 3D
这个数据级别的训练集有一个小BUG,是这个neugasse_station1_xyz_intensity_rgb.7z, 解压之后的名字是station1_xyz_intensity_rgb.txt,自己最好手动去修改成neugasse_station1_xyz_intensity_rgb.txt,不然跑一写程序会存在一些问题
##1. 网址
##2. 下载:
wget -b -i url.txt
对应的url.txt在文末附录
3. 简介
因为我们使用‘户外场景’的数据做实验,这种论文能投的期刊范围会更广一些。比如遥感类的期刊,就很喜欢这种场景,它不喜欢ModelNet40这种只有模型的数据库。
We provide both training and test data as zipped ascii text files with format {x, y, z, intensity, r, g, b}. The ground truth is provided as single column ascii file, where the row ids of the class labeles and the points correspond
如何理解instensity这个通道,可以点击这里或者这里
-
是大范围点云分割的benchmark
-
自然场景超过
40亿
,有ground truth 和hand-labelled,加压后总大小大概150G
-
提供了两种数据:
- semantic-8,包含8类东西(1: man-made terrain, 2: natural terrain, 3: high vegetation, 4: low vegetation, 5: buildings, 6: hard scape, 7: scanning artefacts, 8: cars, 0: unlabeled points)。0lable 没有ground truth也不应该没用于训练。训练集和测试集各
15个
,点云总数超过10亿
, 所以作者怕你的算法受不了,所以提供了一个压缩版本的 - reduce-8训练集和semantic-8一样,测试集只是semantic-8的一部分。测试集是0.01m的均匀下采样.
- semantic-8,包含8类东西(1: man-made terrain, 2: natural terrain, 3: high vegetation, 4: low vegetation, 5: buildings, 6: hard scape, 7: scanning artefacts, 8: cars, 0: unlabeled points)。0lable 没有ground truth也不应该没用于训练。训练集和测试集各
NYUv2
- Kinect 采集
- 1449个稠密的RGBD标注场景
- 3个城市的464个场景
- 407,024 没标注的帧
- 都是用一类或者模块号标注的
NYUv2数据集没有考虑到整个房间的重建,并且通常没有为给定场景提供显著的视点变化。
S3DIS
Stanford large-scale 3D Indoor Spaces Dataset (S3DIS).
- 能够
在线查看点云图
,而且做一些操作
附录
- semantic 3D下载
因为semantic-8和reduce-8的训练集都是一样的,只是测试集不一样而已。
train dataset url.txt
http://www.semantic3d.net/data/point-clouds/training1/bildstein_station1_xyz_intensity_rgb.7z
http://www.semantic3d.net/data/point-clouds/training1/bildstein_station3_xyz_intensity_rgb.7z
http://www.semantic3d.net/data/point-clouds/training1/bildstein_station5_xyz_intensity_rgb.7z
http://www.semantic3d.net/data/point-clouds/training1/domfountain_station1_xyz_intensity_rgb.7z
http://www.semantic3d.net/data/point-clouds/training1/domfountain_station2_xyz_intensity_rgb.7z
http://www.semantic3d.net/data/point-clouds/training1/domfountain_station3_xyz_intensity_rgb.7z
http://www.semantic3d.net/data/point-clouds/training1/neugasse_station1_xyz_intensity_rgb.7z
http://www.semantic3d.net/data/point-clouds/training1/sg27_station1_intensity_rgb.7z
http://www.semantic3d.net/data/point-clouds/training1/sg27_station2_intensity_rgb.7z
http://www.semantic3d.net/data/point-clouds/training1/sg27_station4_intensity_rgb.7z
http://www.semantic3d.net/data/point-clouds/training1/sg27_station5_intensity_rgb.7z
http://www.semantic3d.net/data/point-clouds/training1/sg27_station9_intensity_rgb.7z
http://www.semantic3d.net/data/point-clouds/training1/sg28_station4_intensity_rgb.7z
http://www.semantic3d.net/data/point-clouds/training1/untermaederbrunnen_station1_xyz_intensity_rgb.7z
http://www.semantic3d.net/data/point-clouds/training1/untermaederbrunnen_station3_xyz_intensity_rgb.7z
semancit-8 test
http://www.semantic3d.net/data/point-clouds/testing1/birdfountain_station1_xyz_intensity_rgb.7z
http://www.semantic3d.net/data/point-clouds/testing1/castleblatten_station1_intensity_rgb.7z
http://www.semantic3d.net/data/point-clouds/testing1/castleblatten_station5_xyz_intensity_rgb.7z
http://www.semantic3d.net/data/point-clouds/testing1/marketplacefeldkirch_station1_intensity_rgb.7z
http://www.semantic3d.net/data/point-clouds/testing1/marketplacefeldkirch_station4_intensity_rgb.7z
http://www.semantic3d.net/data/point-clouds/testing1/marketplacefeldkirch_station7_intensity_rgb.7z
http://www.semantic3d.net/data/point-clouds/testing1/sg27_station10_intensity_rgb.7z
http://www.semantic3d.net/data/point-clouds/testing1/sg27_station3_intensity_rgb.7z
http://www.semantic3d.net/data/point-clouds/testing1/sg27_station6_intensity_rgb.7z
http://www.semantic3d.net/data/point-clouds/testing1/sg27_station8_intensity_rgb.7z
http://www.semantic3d.net/data/point-clouds/testing1/sg28_station2_intensity_rgb.7z
http://www.semantic3d.net/data/point-clouds/testing1/sg28_station5_xyz_intensity_rgb.7z
http://www.semantic3d.net/data/point-clouds/testing1/stgallencathedral_station1_intensity_rgb.7z
http://www.semantic3d.net/data/point-clouds/testing1/stgallencathedral_station3_intensity_rgb.7z
http://www.semantic3d.net/data/point-clouds/testing1/stgallencathedral_station6_intensity_rgb.7z
reduce-8 test
http://www.semantic3d.net/data/point-clouds/testing2/MarketplaceFeldkirch_Station4_rgb_intensity-reduced.txt.7z
http://www.semantic3d.net/data/point-clouds/testing2/StGallenCathedral_station6_rgb_intensity-reduced.txt.7z
http://www.semantic3d.net/data/point-clouds/testing2/sg27_station10_rgb_intensity-reduced.txt.7z
http://www.semantic3d.net/data/point-clouds/testing2/sg28_Station2_rgb_intensity-reduced.txt.7z