I have been searching for the solution then I see that there are two ways:

1. Applying the pybullet_ros from this https://github.com/oscar-lima/pybullet_ros, then use ROS to subscribe to the image topic published.

2. Attaching the pybullet's camera to the robot and computing the view matrix at every time step?

My questions are:

- Are these two ways the right things to do, is there any better approach?

- I gravitate towards the latter but I don't know how to do it right yet, Can you give me some solutions?

Thank you very much!!!