MATLAB Answers


Problem using minimal fancoverage in ifanbeam / fan2para

Asked by Carsten Killer on 17 Apr 2015
I have some basic difficulties when using the option FanCoverage='minimal' . In the example below I first apply fanbeam to the phantom image and then attempt to reconstruct the phantom using ifanbeam. It turns out that ifanbeam results in errors for specific combinations of the parameters FSS and FRI.
FSS=1; %FanSensorSpacing
FRI=1; %FanRotationIncrement
D=sqrt(2)*size(I,1); %typical distance for the fan vertex
%perform fanbeam by directly calling the two subfunctions since fanbeam does not support the minimal coverage option
P = radon(I,0:0.5:179.5);
F = para2fan(P,D,'FanSensorSpacing',FSS,'FanRotationIncrement',FRI,...
% perform the inverse operation to reconstruct the phantom
Depending on the choice of FSS and FRI, ifanbeam sometimes gives the following error:
??? Error using ==> fan2para>checkAngles at 266
Internal problem: THETA values must be in the range [min(GAMMA),360) degrees.
Error in ==> fan2para at 193
Error in ==> ifanbeam at 193
[P,oploc,optheta] = fan2para(F,d,...
As an example, the code works for
  • FSS=1, FRI=1
  • FSS=1, FRI=2
  • FSS=1, FRI=3
  • FSS=1, FRI=5
  • FSS=1, FRI=6
  • FSS=1, FRI=10
while the error occurs for the combinations
  • FSS=1, FRI=4
  • FSS=1, FRI=7
  • FSS=1, FRI=8
  • FSS=1, FRI=9
I tried many different combinations, but I really cannot find a reason why some combinations work and others don't. The problem always seems to originate in line 262 of fan2para
isAnyThetaTooSmall = any(thetaDeg < floor(min(gammaDeg)));
At this point I am not able to determine whether the root of the problem lies in the theta-vector or the gamma-vector.
Might there be a problem with the subroutines which build these vectors or am I simply using invalid input parameters that make no sense?


Sign in to comment.

0 Answers