This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

IWR6843ISK-ODS: Overhead 3D People Counting

Part Number: IWR6843ISK-ODS
Other Parts Discussed in Thread: DCA1000EVM, MMWAVEICBOOST, IWR6843

Hello Guys,

Good day.

We have a customer looking to implement the Overhead 3D people counting from the Lab below. They want to capture sensor data via API and visualize it on their website while also visualizing it on monitor screen near the entrances. Is that possible?

They also like to set threshold for people count and ideally want to enable bi-directional counting at entrances to keep track of room occupancy. How can this be done?

Would they need additional hardware such as DCA1000EVM? What are the other hardware and software needed aside from the ones listed on the laboratory?

Overhead 3D People Counting Lab: https://dev.ti.com/tirex/explore/node?node=AF4jhFuBH13mLPnsCXgawg__VLyFKFf__LATEST

Thanks in advance!

Art

  • Hello

    In order to load and run  the demo Just the EVM  IWR6843ISK-ODS  would suffice. It will give the post processed data over uart->USB  port of EVM.

    DCA1000EVM is needed for Raw Data collection used for algorithm development.

    Regards

    Vaibhav

  • Thanks Vaibhav!

    How about the API, will the mmWave SDK be sufficient for that?

    Art

  • Hi Art,

    The UART output of the Overhead People Counting Demo, as explained in the demo user guide, contains the point cloud and track information which is simply visualized by the provided Matlab based visualizer (the source code for which is also provided in the lab). Please look at the "UART Output Data Format" section in the lab user guide for more details. The customer should be able to take this output from UART directly and send it to their visualizer without needing anything other than the mmWave Sensor EVM as listed in the lab user guide.

    Please go through the user guide and get back if you have further questions.

    Regards

    -Nitin

  • Hello Nitin,

    From the solution that you have suggested, they would need an additional PC to connect to sensor board (UART to USB) and a display to the PC to see demo visualization.

    However, they are still unsure how to get the sensor data to their website for analytics and visualization. Also, I apologize for asking again, but I just like to confirm, if they will still be needing DCA1000EVM as well as MMWAVEICBOOST to achieve their purpose as describe above?

    Thanks for looking into this!

    Art

  • Hi Art,

    1. If they can find a way to convert the UART data from mmWave EVM to TCP/IP packets, using a cheaper (perhaps embedded) solution as compared to a PC, they can directly send the sensor data to their website which will be running the visualizer. This way, they won't need the PC.

    2. The DCA1000 EVM is not needed for taking the processed output of the Overhead People Counting demo and visualizing it as described in the customer use case. The complete analog and digital signal processing and tracking algorithm runs on the IWR6843 mmWave sensor and the output is fully processed data containing detected objects and track information. The DCA1000 EVM is needed only if you want to collect the raw ADC sample data from the mnmwave sensor for algorithm development (like we did when developing the overhead people counting demo processing chain).Hopefully that clarifies the doubt but let us know if you need more information.

    Regards

    -Ntiin    

  • Thanks a lot Nitin!

    I'll refer the customer to this post as well in case they have further inquiries with their application.

    I appreciate if you can extend your support to them as well.

    Art