Hello all,
I would like to locate the coordinates of the detected object on the sample image itself and not on the map.
For example: there are 3 detected bounding boxes for 3 objects. I would like to know the exact coordinates of these bounding boxes on this image.
could you please help?
Regards
Please take a look at the tutorial at https://www.nuscenes.org/tutorial
The render_sample_data function can render boxes on the image.
If you then dig into the code you will see that get_sample_data returns the boxes in the coordinate frame of the respective sensor (https://github.com/nutonomy/nuscenes-devkit/blob/master/python-sdk/nuscenes/nuscenes.py#L776).
I have seen the tutorial and I am currently digging into the code, my question is how to get the coordinates of the bounding box related to the image and not to the sensor.
since the origin of the image is the upper left corner, then I would like to get the bounding boxes coordinates related to the upper left corner (virtual origin of my image)
Look at how the boxes are rendered.
The view_points function projects the points to the image: https://github.com/nutonomy/nuscenes-devkit/blob/master/python-sdk/nuscenes/utils/data_classes.py#L568
Thank you for your fast replies. Actually my question is only to view the bounding box coordinates values and not how they are rendered or viewed. just only the x,y,z coordinates of the boxes related to the virtual origin of the image. do you provide that in the NuScenes algorithm?
No. You need to copy the code from that function.