I am trying to use the instance segmentation camera in Carla 0.9.14 to determine what actors I should grab velocity and acceleration data from after each world.tick(). Here is what the docs say:
(https://carla.readthedocs.io/en/latest/ref_sensors/#instance-segmentation-camera)
"The server provides an image with the tag information encoded in the red channel: A pixel with a red value of x belongs to an object with tag x. The green and blue values of the pixel define the object's unique ID. For example a pixel with an 8 bit RGB value of [10, 20, 55] is a vehicle (Semantic tag 10) with a unique instance ID 20-55."
I verified that the cars I am looking at have the correct semantic tag in the red channel, but the values I get on the green and blue channels don't seem to correlate with their 'car.id'.
For a car actor which has a car.id = 317, my RGB vector is [10 (car label), 66, 81]. How do I connect GB values to car.id?
I am trying to use the instance segmentation camera in Carla 0.9.14 to determine what actors I should grab velocity and acceleration data from after each world.tick(). Here is what the docs say:
(https://carla.readthedocs.io/en/latest/ref_sensors/#instance-segmentation-camera) "The server provides an image with the tag information encoded in the red channel: A pixel with a red value of x belongs to an object with tag x. The green and blue values of the pixel define the object's unique ID. For example a pixel with an 8 bit RGB value of [10, 20, 55] is a vehicle (Semantic tag 10) with a unique instance ID 20-55."
I verified that the cars I am looking at have the correct semantic tag in the red channel, but the values I get on the green and blue channels don't seem to correlate with their 'car.id'.
For a car actor which has a car.id = 317, my RGB vector is [10 (car label), 66, 81]. How do I connect GB values to car.id?