Direct Linear Transformation

Objectives

  • Describe the Direct Linear Transformation (DLT)
  • Use Peak Motus to perform a DLT

A procedure for generating three-dimensional data from multiple two-dimensional images.

Direct Linear Transformation
  • For camera 1, we can obtain two equations relating to the three-dimensional object space.
    • Direct Linear Transformation Equations
  • The L's represent camera constants (DLT parameters). In order to find out the values for the camera constants, we must have at least 11 unique equations. If we film enough points in the object space from at least two unique camera views, we can accomplish this.
  • Typically, at least two cameras are used with at least 8 "control points" in the field of view.
    • What is the minimum number of cameras and control points to solve the 11 camera constants?
    • What is the benefit to using additional control points and camera views?
  • Once the camera constants are solved, other objects in the object space can be filmed and three-dimensional coordinates (x, y, z) can be determined.
    • What is the minimum number of cameras to solve for x, y, and z?
    • What is the benefit to using additional camera views?

Mini project #3