• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 2 answers
  • Subscribers 31 subscribers
  • Views 628 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

Original question:

AWR1642BOOST: Use of sensor outside of a car

AWR1642BOOST: Can the range azimuth heatmap see a person behind another person?

Antonios Marios Christonasis
Prodigy 60 points
Part Number: AWR1642BOOST

My question has to do with the range azimuth heatmap that comes from the vehicle occupancy detection demo.

I am trying to use the data coming from the heatmap to count people inside the car as the demo does.

However, in my case, I want to have the sensor outside the window near one of the backseats.

The sensor is looking inside the car through the window.

What I want to know is: If a person is sitting next to that window and another person sits next to him on the other side of the car or even next to him in the middle back seat, will the sensor be able to see the people next to that person?

I have tried to test that and it seems to me that the sensor can detect one person or more vs no person in the back seats, but it cannot differentiate among 1vs2vs3 people sitting in the back seat.

Is there an explanation to that from the nature of the sensor/range azimuth heatmap?

Is the radio wave coming back only from the first person, so there is no signal reaching the people next to him as the sensor sees them on behind another?

The idea is that the sensor has 3 meters range but if a person stands in the way of another person will that affect?

In essence, am I trying to do something that physically cannot be done due to the signal not reaching the people I want to detect?

  • 0
    TI__Guru*** 149314 points

    Yes,

    You will not be able to detect a person behind another person.

    The human body is highly reflective and the radar signal will be completely reflected by the first person

    thank you

    Cesar

  • 0
    TI__Mastermind 21025 points

    >However, in my case, I want to have the sensor outside the window near one of the backseats.

    >The sensor is looking inside the car through the window.

    This is not a good location for occupancy detection, as the angle means that any occupants in the back row will be overlapping at least in the view of the antenna.  This places them closer together in the heatmap, when you want to separate them as much as possible.

    >I have tried to test that and it seems to me that the sensor can detect one person or more vs no person in the back seats, but it cannot differentiate among 1vs2vs3 people sitting in the back seat.

    While our occupancy demo is still evolving, there are third party vendors who have demonstrated that it is quite possible.  Search YouTube for videos.  Most of the time, they start with our chirp and heatmap, then perform their own post-processing to declare occupant detections.

     -dave

  • 0 in reply to db_woodall
    Prodigy 60 points

    Hi Dave and thanks for your answer.

    Do you have any video with successful demonstrations to suggest?

    From the answer of Cesar and also yours, I understand that having the sensor at the side of the car and hence seeing people as they are in a row, will not make it able to count how many people are in the back seat.

    Do you have any suggestions as to where to place the sensor outside of a car to make it possible to count all the people inside it?

    Thanks, Antonios

  • 0 in reply to Cesar
    Prodigy 60 points

    Hi Cesar.

    Do you have any suggestion on where would be the best place to position the sensor on the outside of a car in order to be able to count all the people inside the car from 1 up to 5? I am also doing post-processing on the data, so I want to know the best way (if possible) to position the awr sensor on the outside of a car and based on the range azimuth heatmap data count how many people are sitting inside the car, from 1 up to 5. The use case assumes that I do not own the car, so the sensor cannot be attached to the car. It should be positioned at an external location around 0.5m from the car. Assume for example a parking lot where cars come inside and stay for some time at the entrance bar to get a ticket.

    Thanks, Antonios

  • 0 in reply to Antonios Marios Christonasis
    TI__Mastermind 21025 points

    Yes, here is one:  https://www.youtube.com/watch?v=4KQQWzHzAuo

    I believe the sensor is best mounted along the centerline of the vehicle. The two mounting positions most often considered by vendors are either front mount (around the rear-view mirror area), or roof mount (location depends on the area being monitored).  Mounting the sensor away from the centerline would create the issues already mentioned, and would also increase the likelihood of a position being blocked from view of the sensor.

    If the sensor is mounted to something other than the car, then this is probably not the best algorithm - since the car will likely be moving and there will be limited line of sight to occupants inside the car.