Intel® Distribution of OpenVINO™ Toolkit
Community assistance about the Intel® Distribution of OpenVINO™ toolkit, OpenCV, and all aspects of computer vision-related on Intel® platforms.
6403 Discussions

How to use the --save_detections command line to save the detected bounding box ID?

Lenard-Lim
Beginner
854 Views

Hi, I am using Multi Camera Multi Person Python* Demo ver2020.3 to do the person re-identification.

I use the command line to activate the cameras and re-identify the person:
python3 multi_camera_multi_person_tracking.py \
-i 0 2 \
--m_detector /home/lenard/Downloads/openvino/open_model_zoo/demos/python_demos/multi_camera_multi_person_tracking/person-detection-retail-0013/FP16/person-detection-retail-0013.xml \
--m_reid /home/lenard/Downloads/openvino/open_model_zoo/demos/python_demos/multi_camera_multi_person_tracking/person-reidentification-retail-0103/FP16/person-reidentification-retail-0103.xml \
-d MYRIAD \
--config config.py

But how can I use the --save_detetcions to save the detected bounding box ID? 

Thank you

0 Kudos
8 Replies
Iffa_Intel
Moderator
834 Views

Greetings,


If you refer here: https://docs.openvinotoolkit.org/latest/omz_demos_python_demos_multi_camera_multi_target_tracking_README.html


or if you run the command: python3 multi_camera_multi_person_tracking.py -h

you should be able to see the command options that are available.

If the --save_detections is listed, then you can use it in this particular sample code.


you just need to add --save_detections <outputfilename> as an extra command when you are trying to run the code:

-i 0 2 \

--m_detector /home/lenard/Downloads/openvino/open_model_zoo/demos/python_demos/multi_camera_multi_person_tracking/person-detection-retail-0013/FP16/person-detection-retail-0013.xml \

--m_reid /home/lenard/Downloads/openvino/open_model_zoo/demos/python_demos/multi_camera_multi_person_tracking/person-reidentification-retail-0103/FP16/person-reidentification-retail-0103.xml \

-d MYRIAD \

--config config.py  --save_detections SAVE_DETECTIONS


(the output file would be SAVE_DETECTIONS)


Sincerely,

Iffa


0 Kudos
Lenard-Lim
Beginner
830 Views

Hi Iffa,

Thank you for your solution. May I know the --history_file comand line can use to save the object's ID right?
And may I know do you have any suggestion or reference to add a simple counting system to use this demo counting people?

Regards,
Lenard

0 Kudos
Iffa_Intel
Moderator
827 Views

Generally, the --history_file argument's purpose is to specify the file of dump resulting tracks to a json file from a demo.


You may refer to this source for the question that you ask regarding people counting: https://github.com/intel-iot-devkit/people-counter-cpp


Sincerely,

Iffa


0 Kudos
Lenard-Lim
Beginner
822 Views

Hi Iffa,

Thank your solution. Sorry about I am a beginner to learn computer vision and python now. May I know what is mean by "dump"?

I can simple take the code from the source and add in to this demo. Right?
May I know how this source counting people? Is it when it detect the people then it will count it or need to pass though the line?

Regards,
Lenard

0 Kudos
Iffa_Intel
Moderator
813 Views

Basically dump have the same concept as load.


It's really good for your understanding if you could study the model from the documentation.

This is the model that being used: https://docs.openvinotoolkit.org/latest/omz_models_intel_person_detection_retail_0013_description_person_detection_retail_0013.html


There are also a complete tutorial if you look thoroughly: https://www.youtube.com/watch?v=PxbYqBZgbQE

(this includes deployment)


First step that you should be doing is this: https://docs.openvinotoolkit.org/latest/openvino_docs_install_guides_installing_openvino_linux.html



Ensure the OS is correct accordingly.




Sincerely,

Iffa


0 Kudos
Lenard-Lim
Beginner
810 Views

Hi Iffa,

Some of the part I don't understand. The inputs and outputs part, it just for understand the concept. Right?
Can I use as a parameter to do the counting?

Regards,
Lenard

0 Kudos
Iffa_Intel
Moderator
800 Views

The documentation explains what the model does including the details that are used in Inputs and Outputs.


the model need to be run with the Openvino's sample code. Look carefully how it is done according to the video I gave and also click on the youtube's owner github & read the explaination there.


Sincerely,

Iffa


0 Kudos
Iffa_Intel
Moderator
784 Views

Greetings,


Intel will no longer monitor this thread since we have provided a solution. If you need any additional information from Intel, please submit a new question



Sincerely,

Iffa


0 Kudos
Reply