How to obtain disparity from depth?

[jinan-zhou]

Thank you for sharing this awesome dataset!

I notice that in the paper Realtime Time Synchronized Event-based Stereo, the disparity ground truth of MVSEC is used. Thus, I am wondering how disparity is computed from depth ground truth for this dataset. I know that disparity = focus * baseline / depth. So what are the values of focus and baseline, in the unit of pixels?

[alexzzhu]

You should be able to obtain the focal length and baseline from the provided camera calibration. In general, if you have rectified images, the focal length and baseline can be read from the projection matrix, defined here: http://docs.ros.org/en/melodic/api/sensor_msgs/html/msg/CameraInfo.html