## Download Dataset
Download KITTI 3D object detection data and organize the folders as follows:
dataset/KITTI/
data_object_calib/
training/calib/
testing/calib/
data_object_velodyne/
training/velodyne/
testing/velodyne/
image_2/
training/image_2/
testing/image_2/
label/
training/label2/
training/
val/ss3d_results/
val/fpnet_val_results/
testing/rrc_results/
mkdir dataset_generate/
prepare_trainval_dataset_gt2dbbox.py
split_data_trainval.py
image_sets/ss3d_train.txt
image_sets/ss3d_val.txt
image_sets/ss3d_val.txt
image_sets/fpnet_train.txt
image_sets/fpnet_val.txt
image_sets/ss3d_trainval.txt
image_sets/ss3d_test.txt
## Usage
### Prepare Frustum pointcloud
config kitti dataset path in prepare_trainval_dataset_gt2dbbox.py
```
python prepare_trainval_dataset_gt2dbbox.py
```
config list file path in split_data_trainval.py
```
python split_data_trainval.py
```
### Prepare PointCNN .h5 file
python prepare_kitti_partseg_data.py -f ../dataset_root/ -d 9
基于点云的PointCNN点卷积神经网络，使用KITTI数据集对车辆的点云输入进行语义分割Python源码+源代码+文档说明
版权申诉
76 浏览量
2024-01-11
22:02:23
上传
评论
收藏 108KB ZIP 举报
PointCNN-self-driving-cars-master.zip （23个子文件） 
PointCNN-self-driving-cars-master data_conversions kitti_velodyne_prepare image_sets 
fpnet_val.txt 26KB ss3d_train.txt 23KB fpnet_trainval.txt 51KB fpnet_train.txt 25KB ss3d_val.txt 26KB fpnet_test.txt 51KB
prepare_trainval_dataset_gt2dbbox.py 13KB split_data_trainval.py 6KB prepare_kitti_partseg_data.py 10KB README.md 1KB pointcnn_kitti.py 14KB train_val_seg_kitti.py 23KB pointcnn_seg_kitti.py 379B pointcnn_seg_kitti train_val.sh 827B test.sh 1KB kitti_x4_2048_xyrgbi.py 3KB __pycache__ kitti_x4_2048_xyrgbi.cpython-35.pyc 2KB pointfly.py 17KB eval_vis kitti_seg_merge_vis.py 19KB kitti_seg_bbox_merge_vis.py 23KB README.md 619B data_utils.py 8KB test_seg_kitti.py 8KB资源评论
