15/11/2004


___________________________________________
Asymmetry

-The problem was solved.
-Correct inverse warp was needed.
-See: http://www2.cs.man.ac.uk/~schestr0/Temp2/Exp/asym.zip
-Implications:
     o Reconstruction from the model finally works properly.
     o The objective function is more correctly defined and the results appear to have improved.
     o Calculation of model discrepancy is done correctly.


___________________________________________
Model-Building

-Perturbation of initial warps to allow inclusion of the shape component.
-Height problem appears less crucial when shape is accounted for.
-The function is not stable at the correct solution, even with proper shape-intensity weighting.


___________________________________________
Videos

-Videos of models and registration were generated.

-Model animation:
     o A pixel representation and plot representation of shape, intensity and combined model.
     o Video compression under Windows is possible now; normal file compression under UNIX.
     o See: http://www2.cs.man.ac.uk/~schestr0/Temp2/Exp/Videos.zip;
               show_models_before_reg-502-256-combined_model_at_end.avi
     o Other representation is in the same ZIP archive.

-Registration:
     o See: http://www2.cs.man.ac.uk/~schestr0/Temp2/Exp/Videos.zip;
              show_mesh_Attempt_4-1.avi


___________________________________________
Integration in the OF

-Implementation is in progress.
-Carole is heavily involved.


___________________________________________
Other Items (implementation)

-Mesh rotation has been corrected to generate a more consistent viewpoint in videos.
-Tom Hartley(RF from UCL) - request for a documentation tool (see E-mail from Nov. 12th).


___________________________________________
Organisational

-Form 10 was amended and handed in.
-UCL/IC visit