• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 31 subscribers
  • Views 431 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.

AWR1642: radar camera calibration and radar axis issues.

DEEPAK MURALI
Prodigy 50 points
Part Number: AWR1642

Hi,
I am using TI mmwave radar AWR1642 in short range. Through data collection, I have observed that the radar provides detection up-to 10m range in depth(X axis). Another feature i observed is that the axis of radar is aligned so that the origin is at centre and x axis increases from zero to max, y axis ranges from positive max to negative max. While calibrating radar and camera

  1. should i shift the origin towards the side?
  2. what is the maximum detection range horizontally(Y axis)?

Note: On the image below,Image from, it shows that horizontal range is -8.5 to 8.5m that is approx 17m. Is this valid ?

Kindly help us to complete the radar camera calibration, if you feel that i am doing a wrong way to do radar camera calibration.

Thanks in advance.
Deepak

  • 0
    TI__Guru 58035 points

    Hello Deepak,

    Radar would provide you with the range and angle information for each object. With this you can estimate the X and Y distance of the object from the radar sensor. The radar sensor would be the origin in this case.

    I am not sure what kind of calibration you are doing between the radar and camera sensors.

    regards,
    vivek

  • 0 in reply to Vivek Dham
    Prodigy 50 points

    , Thank you for your kind response. I have already converted the polar to Cartesian coordinate.  i would like to know how to project x,y coordinates of the radar onto the 2d image of the camera. Presently i am using usb camera and i did camera calibration. I have tried two methods to do the radar to camera calibration. First one is CV2.SolvePNP and second method is homography. Unfortunately, the both experiment results were not accurate. Its RMSE values are higher even after collecting more than 30 corresponding points. I have tried to reduce RMSE value, but its not working. Could you suggest any solution to implement the radar to camera calibration ?

  • 0 in reply to DEEPAK MURALI
    TI__Guru 58035 points

    Hello Deepak,

    Unfortunately, we dont have experience in the camera fusion. You could possibly get more details online with some standard literature or publications on sensor fusion. 

    Regards,

    vivek