In order to correctly project images onto televisions,
            bulletin boards, and other quadrilaterals in an image sequence use
            the 4-Point tracking workflow.
         
         With four-point tracking
            (also known as corner pinning), you use four trackers to generate
            tracking data for anchoring the four corners of a bilinear surface
            to background footage.
         
         The reference points
            you use must be well-defined; it is recommended that you plan them
            when shooting the sequence (for example, add markers to the scene).
            It is not always possible to do four-point tracking when the reference points
            are not well-defined.
         
         In the following example,
            the goal is to pin an image of a jet to the screen of a monitor
            as the camera moves.
         
         NoteYou can also use
               the Reaction tool, and the position and translation parameter controls
               of the Warp2D tool, and the position and translation vertices parameter
               controls of the Garbage Mask, and Remove Dust tools.
            
         To corner pin and track a bilinear surface
            to background footage:
         
         
            - In the Schematic view, the basic dependency
               graph has been set up with the following nodes:
                  
                     
                        
                        
                        
                           | Node | Purpose |  
                           | Input image (front) | This is the footage that will tracked. |  
                           | Input image (back) | This image will be corner-pinned onto the front footage. |  
                           | Keyer | The Keyer is used to remove the blue portion of the monitor. |  
                           | Tracker super tool | The Tracker will track and analyze the movement of the
                              four corners of the monitor. The Tracker super tool is used because
                              multiple Tracker Analyzers are required. |  
                           | Blend & Comp | The Blend & Comp node will combine the two processed
                              images. |  
                           | 2D Transform | The 2D transform tool will apply the necessary transformations
                              to the image to be corner-pinned. |  
                           | Output | Outputs the final composition. |  
 
 
- Remove the blue screen from the monitor
               with the Keyer.
            
- Create four Tracker Analyzers and name
               them as per their corner positions, then position them in the locations
               that will make up the four corners to pin.
            
- Adjust the Display, Analyze, Reference
               box and Tracking box settings.
            
- Analyze each track (do not forget to
               reset the footage back to the start frame after each analysis).
               You can also select the first Tracker Analyzer and Shift + click the last one to select
               all trackers to analyze at the same time.
            
- Connect the RGBA output from the Keyer
               to the Front input of the Blend & Comp node, and the output
               of the second image into the Back input of the Blend & Comp
               node.
            
- Add a 2D Transform tool to the graph
               between the second image and the Blend & Comp node.
            
- With the 2D Transform tool highlighted,
               select 4 Point from the Transform Type menu.
            
- Click Fit To Source
            
- Right-click on the Destination label
               and select Set Trackers.
            
- Select the trackers in the same sequence
               as you created them from the Tracker Selector window and activate
               the Use Offset toggle.
            
- Click Link.
               The bilinear image is
                  pinned to the background image.
                
- If the corner-pinned image is too small
               or too large for the screen it's replacing, create another transform
               and adjust its scale while in the SRT transform type.