Fly4PET JavaScript Demo





Iterative paradigm for tomography reconstruction
(click here to view the flowchart in full resolution)
  1. Load Measured projections
  2. Set the position of flies, i.e. create initial guess
  3. Compute projections from the current reconstruction
  4. Compute the population's perfomance, i.e. the global fitness
  5. Select a random fly (fly_to_kill)
  6. Remove the fly's contribution
  7. Compute the population's performance without the selected fly
  8. Compare the performances, i.e. compute the fly's local fitness
  9. If the local fitness is equal to or greater than the selection threshold,
    then go to Step 10,
    else go to Step 11
  10. Restore the fly's contribution, then go to Step 5
  11. Select a genetic operator
  12. If the genetic operator is mutation,
    then go to Step 13,
    else go to Step 19
  13. Select a random fly (fly_to_reproduce)
  14. Remove the fly's contribution
  15. Compute the population's performance without the selected fly
  16. Compare the performances, i.e. compute the fly's local fitness
  17. Restore the fly's contribution'
  18. If the local fitness is smaller than the selection threshold,
    then go to Step 13,
    else go to Step 20
  19. Replace fly_to_kill by a new fly with a random position, go to Step 22
  20. Replace fly_to_kill by a new fly based on fly_to_reproduce
  21. Add the fly's contribution to the population
  22. If stop the reconstruction,
    then go to Step 23,
    else go to Step 4
  23. Build problem answer from final population