In this video, we learn how to convert pixels (camera units) to centimeters (world coordinates) by measuring the field of view of the camera.
In this video, we learn how to convert an object position (in units of centimeters) from camera frame coordinates to the manipulator base-frame coordinates. We do this by finding the homogeneous transformation matrix from the base-frame to the camera frame, then multiplying it by the position in the camera frame.
To complete this lab activity, make a video that includes the following in one video:
(1) You saying your name (2) The output of your Python code while you place the nut in different locations on the board. Show that the Python code correctly reports the location of the nut in the base frame X and Y, in units of centimeters, with less than 1cm of error in each direction.
Graduate students also have to do the following: (3) Place your camera looking at your board at an angle, and use Excel to create a mapping between camera coordinates and world coordinates. Show that if you enter an X,Y position in camera coordinates (pixels), your Excel sheet correctly indicates the X,Y position in world coordinates (centimeters).
In this video, we will look at the case where the camera field of view is not parallel to the camera sensor. In other words, the camera is 'tilted' relative to the world. We will use a 'mapping' approach to figure out the transformation between camera coordinates and world coordinates. This is the type of approach taken by widely-used 'structured light' sensors.