Layer rgb not found, adding it to the semantic map

[MRZHUGH]

[ INFO] [1709118694.042549165]: [ElevationMappingCupy] finish initialization
Layer rgb not found, adding it to the semantic map
[elevation_mapping-1] process has died [pid 237549, exit code -11, cmd /emcu_ws/devel/lib/elevation_mapping_cupy/elevation_mapping_node __name:=elevation_mapping __log:=/.ros/log/96750bb0-d626-11ee-a708-5f7e327b3b63/elevation_mapping-1.log].
log file: ~/.ros/log/96750bb0-d626-11ee-a708-5f7e327b3b63/elevation_mapping-1*.log
all processes on machine have died, roslaunch will exit

After print ‘Layer rgb not found, adding it to the semantic map’ , all processes on machine have died.
Can you help me to solve this question? thanks!

[mktk1117]

Oh, this might be the input image issue.
What kind of image did you subscribe? Was it a color image?

[MRZHUGH]

Oh, this might be the input image issue. What kind of image did you subscribe? Was it a color image?

subscribers:
front_cam:
# topic_name: '/camera/depth/points'
topic_name: '/camera/depth/color/points'
data_type: pointcloud
color_cam: # for color camera
topic_name: '/camera/color/image_raw'
camera_info_topic_name: '/camera/color/camera_info'
data_type: image
semantic_cam: # for semantic images
topic_name: '/front_cam/semantic_image'
camera_info_topic_name: '/front_cam/camera/depth/camera_info_resized'
channel_info_topic_name: '/front_cam/channel_info'
data_type: image

As above, when I use the D435i camera, make sure to input color pictures.

[MRZHUGH]

[ INFO] [1709173219.924276172]: Subscribed to Image topic: /camera/color/image_raw, Camera info topic: /camera/color/camera_info. Channel info topic: Not found. Using channels: rgb
[ INFO] [1709173219.925903082]: Subscribed to PointCloud2 topic: /camera/depth/color/points
[ INFO] [1709173219.928238252]: Subscribed to Image topic: /front_cam/semantic_image, Camera info topic: /front_cam/camera/depth/camera_info_resized, Channel info topic: /front_cam/channel_info
[ INFO] [1709173220.260316890]: pointcloud_channel_fusion_dict: {'default': 'average', 'rgb': 'color'}
[ INFO] [1709173220.260602523]: image_channel_fusion_dict: {'default': 'exponential', 'feat_.*': 'exponential', 'rgb': 'color'}
Start loading plugins...
Loaded plugins are min_filter smooth_filter inpainting erosion
[ INFO] [1709173222.165785336]: [ElevationMappingCupy] finish initialization
Layer rgb not found, adding it to the semantic map
[elevation_mapping-1] process has died [pid 256269, exit code -11, cmd

As above, terminal outputs all information.

[MRZHUGH]

RUN roslaunch elevation_mapping_cupy turtlesim_semantic_pointcloud_example.launch, also have the question.

image_topic: /zed2i/zed_node/left/image_rect_color
pub_all: false
publish_feature_image: false
publish_segmentation_image: false
segmentation_image_topic: /semantic_sensor/sem_seg
segmentation_model: lraspp_mobilenet_v3_large
semantic_segmentation: false
sensor_name: front_cam
show_label_legend: false
topic_name: /elvation_mapping/pointcloud_semantic

--------------End of Parameters-----------------------
[('x', <class 'numpy.float32'>), ('y', <class 'numpy.float32'>), ('z', <class 'numpy.float32'>), ('rgb', <class 'numpy.float32'>), ('person', <class 'numpy.float32'>), ('grass', <class 'numpy.float32'>), ('tree', <class 'numpy.float32'>), ('max', <class 'numpy.float32'>)]
Start loading plugins...
[spawn_urdf-3] process has finished cleanly
log file: ~/.ros/log/3b31af8c-d6db-11ee-a901-8dcfa0cb3c92/spawn_urdf-3*.log
Loaded plugins are min_filter smooth_filter inpainting semantic_filter features_pca
[ INFO] [1709194760.818904178, 0.631000000]: [ElevationMappingCupy] finish initialization
Layer rgb not found, adding it to the semantic map
[elevation_mapping-7] process has died [pid 29444, exit code -11,

[mktk1117]

Oh so it does happen with the turtlesim example?
Maybe there's some issue in the dependencies.

Have you tried with docker already?

[MRZHUGH]

Oh so it does happen with the turtlesim example? Maybe there's some issue in the dependencies.

Have you tried with docker already?

Thank you for your reply. The main reason is that there is no detailed error message and I don’t know how to fix it. I am currently trying to run it under docker.

[MRZHUGH]

Oh so it does happen with the turtlesim example? Maybe there's some issue in the dependencies.

Have you tried with docker already?

I only positioned the channel input in this function to only have rgb.

def update_layers_pointcloud(self, points_all, channels, R, t, elevation_map):

Furthermore, I'm not sure if my parameters are misconfigured：

[mktk1117]

Thanks for sharing the details.
Could you check what's in the /camera/camera_info' topic?

If it is a color image, you can leave it, or add ["rgb"] as the channel name.