Skip to content
This repository has been archived by the owner on Aug 7, 2024. It is now read-only.

Run with lates librealsense and ros image topic #173

Open
Lilienthal-GF opened this issue May 11, 2021 · 3 comments
Open

Run with lates librealsense and ros image topic #173

Lilienthal-GF opened this issue May 11, 2021 · 3 comments

Comments

@Lilienthal-GF
Copy link

Hello everyone,
i am having trouble getting the ros_openvino_toolkit package to run with the dev-ov2020.3 branch.

The platform is built on Ubuntu 18.04.5 LTS with the 5.4 Kernel. Therefore i would like to run the lates version of librealsense 2.45.0 and the corresponding realsense-ros package 2.3.0.

I installed the openvino 2020.3 release and confirmed the demo applications on the setup and i can succesfully run catkin_make on the workspace. But if i launch the

roslaunch vino_launch pipeline_object_topic.launch

the realsense camera gets launched correctly (i can view the image in the rostopic), yet i get the error message

[ INFO ] Parsing InputDvice: RealSenseCameraTopic
[ ERROR ] Invalid input device name: RealSenseCameraTopic
[ ERROR ] currently one pipeline only supports ONE input.

I traced this back to the pipeline_manager.cpp and discovered that the RealSenseCameraTopic is commented out in the dev-ov2020.3 branch of ros_openvino_toolkit. Is there a way to fix this?

@Corsair-cxs
Copy link
Contributor

Thank you for your reportation!
I've tried to fix the bug in #174 , could you please test the new pr it in your computer?

Best Wishes

@Lilienthal-GF
Copy link
Author

I can confirm that merging the commit fixes the problem. I changed the pipeline_object_topic.launch to launch the rs_camera-node again. But now i the following error message when i launch the pipeline_object_topic.launch

[ ERROR ] Unexpected CNNNetwork format: it was converted to deprecated format prior plugin's call
[pipeline_with_params-4] process has died [pid 17657, exit code 1

I downloaded and converted the mobile-ssd from the open_model_zoo in the openvino install folder, i.e.

sudo python3 ~/intel/openvino/deployment_tools/model_optimizer/mo.py --input_model ~/intel/openvino/deployment_tools/open_model_zoo/tools/downloader/public/mobilenet-ssd/mobilenet-ssd.caffemodel --output_dir /opt/openvino_toolkit/models/object_detection/mobilenet-ssd/caffe/output/FP16 --data_type=FP16 --mean_values [127.5,127.5,127.5] --scale_values [127.5]

Do i need to use a different tool to download or convert the net?

@LewisLiuPub
Copy link
Contributor

I think you need to check the version of OpenVINO Toolkit you installed. branch dev-ov2020.3 is now the latest stable branch, you also need to use the same version (2020.3LTS) of Intel OpenVINO toolkit (including Inference Engine and Model Optimizer).
Please try and let us know your results. @Lilienthal-GF

Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants