对原始点云进行缩放重构,确保能够将点云放进open3d的视窗下
input_txt_dir = r"C:\Users\Administrator\Desktop\mp4\other_input\input"
output_dir = r"C:\Users\Administrator\Desktop\mp4\other_input\combine_txt"
output_mp4_dir = rf"C:\Users\Administrator\Desktop\mp4\other_output"
input_txt_dir
为输入点云的目录;
output_dir
为重构点云的输出目录;
output_mp4_dir
为mp4视频的输出目录。
需要在命令行中能识别到ffmpeg即可。
conda create -n semantic_rotation python==3.8 -y
conda activate semantic_rotation
pip install open3d==0.17.0
pip install opencv-python==4.10.0.82
git clone https://github.com/Timothy7890/semantic_rotation.git
cd semantic_rotation
pip install -e .
conda activate your_env
git clone https://github.com/Timothy7890/semantic_rotation.git
cd semantic_rotation
pip install -e .
举例1
python .\PipelineRun_Sematic_R_cmd.py --input_dir "C:\Users\Administrator\Desktop\mp4\other_input\input" --output_dir "C:\Users\Administrator\Desktop\mp4\other_input\combine_txt" --output_mp4_dir "C:\Users\Administrator\Desktop\mp4\other_output" --overwrite
举例2
python .\Pipeline\Run_Sematic_R_cmd.py --input_dir "E:\0_Code\python\semantic_rotation\data\input" --output_dir "E:\0_Code\python\semantic_rotation\data\combine_txt" --output_mp4_dir "E:\0_Code\python\semantic_rotation\data\output" --overwrite
- txt格式;
- 间隔符为空格;
- 前七列是x y z r g b 和语义信息。
#输出的视频时长为14s,不支持自定义调节
#在第七秒进行过渡,过渡时长为0.5s
#设置y轴为点云旋转轴