##### System information (version)
- OpenVINO => 2021
- Operating System / Platform => Windows 64 Bit
- Compiler => Anaconda command prompt
- Problem classification: Multi camera multi person tracking
I used this program based on openvino demo multiple-camera_multiple-people tracking to study computer vision, but I can't make it work, neither apply to other videos. It really works detecting people using one camera, but it does not work when I use more cameras. The program duplicates the boxes created based on the first camera on the other ones, without matching people on different videos. Do I need to configure the position of the cameras? I tested using the videos and basic command to run of the repository.
Also, it is not resizing the video properly to show.
##### Steps to reproduce
using command prompt
go to openvino/bin and execute setupvars.bat
go to the repository folder and run
python multi_camera_multi_person_tracking.py -i datasets/HallWayTracking/videos/001.avi datasets/HallWayTracking/videos/002.avi datasets/HallWayTracking/videos/005.avi -m model/intel/person-detection-retail-0013/FP32-INT8/person-detection-retail-0013.xml --m_reid model/intel/person-reidentification-retail-0031/FP32-INT8/person-reidentification-retail-0031.xml --config config.py -d GPU
Thank you for reaching out to us. I would suggest you refer to the following links as your references for the 'Multi-Camera Multi-Target Tracking Python* Demo' sample:
The inaccurate result obtained might be due to deprecated Inference Engine classes and hardware limitations. Please try and use 'CPU' as the device to run the demo instead of 'GPU'. As a side note, you can choose and download sample videos through this link: https://github.com/intel-iot-devkit/sample-videos. The sample videos can be used as demo inputs.
This thread will no longer be monitored since we have provided a solution. If you need any additional information from Intel, please submit a new question.