parser = argparse.ArgumentParser(description='Inverse Radon Transform -- FBP',

formatter_class=argparse.ArgumentDefaultsHelpFormatter)

parser.add_argument('-Di', '--pathin', dest='pathin', default='./',

help='Specify path to input data')

parser.add_argument('-i', '--sino', dest='sino',

help='Specify name of input reco')

parser.add_argument('-Do', '--pathout', dest='pathout',

help='Specify path to output data')

parser.add_argument('-o', '--reco', dest='reco',

help='Specify name of output reconstruction')

parser.add_argument('-g', '--geometry', dest='geometry',default='0',

help='Specify projection geometry; @@@@@@@@@@@@@@@@@@@@@@@'

+' -g 0 --> equiangular projections between 0 and 180 degrees (default);'

+' @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@'

+' -g angles.txt use a list of angles (in degrees) saved in a text file')

parser.add_argument('-c',dest='ctr', type=myFloat,

help='Specify the center of rotation ( default = the middle pixel of the image )')

parser.add_argument('-k',dest='niter', type=int, default=3,

help='Choose number of iterations')

parser.add_argument('-z',dest='edgepad', type=myFloat, default = 0.0,

help='Apply edge padding')

parser.add_argument('-p',dest='plot', action='store_true',

help='Display check-plots during the run of the code')

args = parser.parse_args()

## Exit of the program in case the compulsory arguments,

## are not specified

if args.sino is None:

parser.print_help()

print('ERROR: Input sino name not specified!')

sys.exit()

nang = sino.shape[0]

npix_old = sino.shape[1]

npix_pad = int( padding * npix_old )

print("Padded pixels on one side = ", npix_pad)

columnStart = np.array( sino[:,0] ).reshape( nang , 1 )

columnEnd = np.array( sino[:,npix_old-1] ).reshape( nang , 1 )

sino_aux1 = np.ones(( 1 , npix_pad )) * columnStart

sino_aux2 = np.ones(( 1 , npix_pad )) * columnEnd

sino_pad = np.concatenate(( sino_aux1 , sino , sino_aux2 ), axis=1 )

fig = plt.figure()

ax = fig.add_subplot(111)

ax.imshow( reco , origin="lower" , cmap = cm.Greys_r ,

interpolation='nearest' )

plt.title( title )

numrows, numcols = reco.shape

def format_coord(x, y):

col = int(x+0.5)

row = int(y+0.5)

if col>=0 and col<numcols and row>=0 and row<numrows:

z = reco[row,col]

return 'x=%1.4f, y=%1.4f, value=%1.4f'%(x, y, z)

else:

return 'x=%1.4f, y=%1.4f'%(x, y)

ax.format_coord = format_coord

if args.pathout is None:

pathout = args.pathin

else:

pathout = args.pathout

if pathout[len(pathout)-1] != '/':

pathout += '/'

if args.reco is None:

filename = args.sino

filename = filename[:len(filename)-4]

filename += '_rec_isart.DMP'

filename = pathout + filename

else:

filename = pathout + args.reco

## Initial print

print('\n')

print('########################################')

print('############# PY-SART #############')

print('########################################')

print('\n')

## Get the startimg time of the reconstruction

startTime = time.time()

## Get input arguments

args = getArgs()

## Get path to input reco

pathin = args.pathin

if pathin[len(pathin)-1] != '/':

pathin += '/'

## Get input reco

## You assume the reco to be square

sino_name = pathin + args.sino

sino = io.readImage( sino_name )

npix = sino.shape[1]

nang = sino.shape[0]

print('\nInput sino:\n', sino_name)

print('Number of projection angles: ', nang)

print('Number of pixels: ', npix)

## Show reco

if args.plot is True:

checkPlot( sino , 'Sinogram' )

## Get projection geometry

## 1) Case of equiangular projections distributed in [0,180)

if args.geometry == '0':

angles = np.arange( nang )

angles = ( angles * 180.0 )/myFloat( nang )

print('\nDealing with equally angularly spaced projections in [0,180)')

## 2) Case of list of projection angles in degrees

else:

geometryfile = pathin + args.geometry

angles = np.fromfile( geometryfile , sep="\t" )

nang = len( angles )

print('\nReading list of projection angles: ', geometryfile)

print('Number of projection angles: ', nang)

## Get center of rotation

if args.ctr is None:

ctr = npix * 0.5

else:

ctr = args.ctr

print('\nCenter of rotation placed at pixel: ', ctr)

## Apply edge padding

if args.edgepad != 0:

npix_old = npix

sino = edgepad( sino , args.edgepad )

sino = sino.astype( myFloat )

npix = sino.shape[1]

print('\nEdge padded sinogram size: ', sino.shape[0],' X ', sino.shape[1])

## Compute iradon transform

print('\nPerforming SART iradon ....\n')

niter = args.niter

print('\nNumber of iterations:', niter)

print('\nIter number 1 ....')

reco = iradon_sart( np.rot90( sino ).astype(np.float64) ,

angles.astype(np.float64) )

for i in range( niter - 1 ):

print('\nIter number ', i +2,' ....')

reco[:,:] = iradon_sart( np.rot90( sino ).astype(np.float64) ,

angles.astype(np.float64) ,

image=reco )

## Crop image

if args.edgepad:

i1 = int( 0.5 * ( npix - npix_old ))

i2 = i1 + npix_old

reco = reco[i1:i2,i1:i2]

## Show sino

if args.plot is True:

checkPlot( reco , 'SART reconstruction' )

## Save sino

saveReco( reco , args )

## Time elapsed for the computation of the radon transform

endTime = time.time()

print('\n\nTime elapsed: ', (endTime-startTime)/60.0,' min')