Height map Conversion

Given a point set P^camera retrieved by the camera

where each point is represented as

which includes both position w.r.t. camera coordinate and the color information.

With the transformation matrix obtained by hand-eye calibration, we can transform the point set into arm coordinate, expressed as P^arm
We defines the workspace in arm coordinate

and gets the subset of P^arm, denoted as P^workspace

For each point in P^workspace, we can compute the re-projected index with

and set the pixel at (u, v) as (r, g, b) and z-z_min for color and depth height-map, respectively, and build the mapping between (u, v) and
(x, y, z).

For pixel that does not contain a corresponding point, we simply leaves black (i.e., r=g=b=0) and zero height for color and depth height-map.

The conversion is illustrated as the following

• Related Code

  • Get P^{workspace}
  • Convert Point Cloud to Images