docker image error

[zhangyuran-gg]

我使用的是window11和docker_desktop，使用镜像humble/noetic时出现了这些问题.
bug1:

PS C:\Users\zyr> docker run --rm -v D:\data\bags:/tmp/input_bags -v D:\data\result:/tmp/preprocessed koide3/direct_visual_lidar_calibration:humble ros2 run direct_visual_lidar_calibration find_matches_superglue.py /tmp/preprocessed --rotate_camera 90
Traceback (most recent call last):
  File "/root/ros2_ws/install/direct_visual_lidar_calibration/lib/direct_visual_lidar_calibration/find_matches_superglue.py", line 10, in <module>
    import torch
ModuleNotFoundErr

bug2:

docker run --rm -v D:\data\bags:/tmp/input_bags -v D:\data\result:/tmp/preprocessed koide3/direct_visual_lidar_calibration:humble ros2 run direct_visual_lidar_calibration initial_guess_manual /tmp/preprocessed
camera_model:plumb_bob
intrinsics  :4
distortion  :5
loading /tmp/preprocessed/cam_lidar1.bag.(png|ply)
image_size:1920x1080
camera_fov:84.9796[deg]
glfw error 65544: X11: The DISPLAY environment variable is missing
failed to initial

能帮我看一下吗

[VeeranjaneyuluToka]

You will have to use the image with GUI as below as mentioned in docker.md file
docker run
--rm
--net host
--gpus all
-e DISPLAY=$DISPLAY
-v $HOME/.Xauthority:/root/.Xauthority
-v /path/to/input/bags:/tmp/input_bags
-v /path/to/save/result:/tmp/preprocessed
koide3/direct_visual_lidar_calibration:humble
ros2 run direct_visual_lidar_calibration preprocess -a -v /tmp/input_bags /tmp/preprocessed