NVIDIA-ISAAC-ROS / isaac_ros_pose_estimation

Deep learned, NVIDIA-accelerated 3D object pose estimation
https://developer.nvidia.com/isaac-ros-gems
Apache License 2.0
142 stars 21 forks source link

How do I perform realtime isaac_ros_centerpose using an Intel D435 camera? #33

Closed cobang0111 closed 6 months ago

cobang0111 commented 7 months ago

Hello, Sorry, I am a complete beginner.. Thanks to your repository, I have been able to follow along and study well.

I am truly grateful. Currently, I have succeeded in executing the isaac_ros_centerpose Quickstart and performing the /image topic with centerpose, and I am in the process of analyzing this in real-time using the Realsense D435 camera. Screenshot from 2024-02-15 13-40-08

I have finished installing the Realsense camera inside the Docker environment, and I can now confirm it in rviz2 as shown in the following picture. Screenshot from 2024-02-15 14-45-32 Screenshot from 2024-02-15 14-47-25

Therefore, I modified the source as below and executed several commands. from centerpose_visualizer_node.cpp

only modified

VIDEO_INPUT_TOPIC_NAME

CAMERA_INFO_INPUT_TOPIC_NAME like below Screenshot from 2024-02-15 17-54-20

and from isaac_ros_centerpose_triton.launch.py

only modified

default_value of input_image_width, and input_image_height image

1st terminal

ros2 launch realsense2_camera rs_launch.py depth_module.profile:=640x480x30 pointcloud.enable:=true rgb_camera.color_format:=bgr8 image

2nd terminal

ros2 launch isaac_ros_centerpose isaac_ros_centerpose_triton.launch.py model_name:=centerpose_shoe model_repository_paths:=['/tmp/models']

3rd terminal

ros2 topic list image

when ros2 run rqt_image_view rqt_image_view

/centerpose/image_visualized image

/image also nothing appeared

/camera/camera/color/image_raw realtime video of Intel realsense D435 image

I'm not sure how I should modify it so that the analysis is based on /camera/camera/color/image_raw. Thank you in advance.

cobang0111 commented 6 months ago

image I solved this problem after modified the name of topic ros2 run topic_tools relay /camera/camera/color/image_raw /image ros2 run topic_tools relay /camera/camera/depth/camera_info /camera_info