How to obtain a high precision depth map

What do the numbers in the array information printed out by the depth map mean? The array I printed out is all integers, ranging from 0 to 11, which feels wrong. Does the number represent the distance between the camera and the object? If so, what is its unit? Is there a good way to get a high accuracy depth map?

To get the depth map values you should use retrieveMeasure function, I suspect you displayed value from the normalized depth map image for display (retrieveImage).

The unit can be set with InitParameters::coordinate_units passed to the Camera::open function. The default is millimeters.

Ok, thank you. I want to merge the RGB image with the depth map of the image to make a new image. Now the question is how can the depth map be saved with high precision? Is there a code to save both the depth information and RGB information of a frame directly from the video?