chessfftomo_process_changes.py

import tifffile
import matplotlib.pyplot as plt
import os
import numpy as np
from scipy.ndimage.interpolation import shift
import tomopy
import scipy.ndimage
import skimage.measure
import cv2

def chess_fileio_proj(datapath, prefix, fnumstart, fnumend, imgsize = [2048, 2048],
                          print_infile = False, print_fcounter = False):
    '''
    load frames from datapath with prefix in order
    '''    
    
    numframes = fnumend - fnumstart + 1
    frames = np.zeros([numframes, imgsize[0], imgsize[1]])
    for fnum in range(fnumstart, fnumend+1):
        infile = os.path.join(datapath,prefix+'{0:05d}'.format(fnum)+'.tif')
        if print_infile is True:
            print(infile)
        frames[fnum-fnumstart, :, :] = tifffile.imread(infile)
        
            
    frames = np.array(frames)
    print('    loaded all projects.')

    return frames

def chess_proj(chess_datapath, fname, imgsize = [664, 590]):

    '''
    Load image stacking 
    '''
    image_files = chess_datapath # Your list of files
    image_height = imgsize[0] 
    image_width = imgsize[1]
    image_stack = np.empty((image_height, image_width, 1441)) # Create empty HxWxN array/matrix

    for i, fname in enumerate(image_files):
         #img = cv2.imread(fname, cv2.CV_LOAD_IMAGE_GRAYSCALE)
         img = cv2.imread(fname, cv2.IMREAD_GRAYSCALE)
         image_stack[:,:,i] = img # Set the i:th slice to this image

    proj = image_stack

    return proj
    
def srxfftomo_findcenter(corrected_proj_tiff = None, proj = None, 
                            save_find_center = True, autocheck = False,
                            check_cen_range_step = [390, 410, 1], 
                            auto_selecslice = True, cen_slice = 400,
                            outpath = None, samplename = None):

    '''
    input needs to be either 
    1) corrected_proj_tiff, provided as a file path and name, 
    e.g. '/home/xf05id1/localdata/TomoCommissioning/testsample_01/testsample_01_corrected.tiff'
    it is a stack of tiff generated by srxfftomo_correction, where background and stage round out have been corrected
    2) projection numpy array, returned from srxfftomo_correction
    
    The 'proj' array assignment has priority. If it is None, proj was be read from corrected_proj_tiff 
    
    input example:
        check_cen_range_step = [390, 410, 1]
        cen_slice = 400    
    
    return: 
    proj (numpy array, float), with negative natural log has been taken
    
    '''    
    if proj is None:
        print('proj is not asigned, load corrected projection from:', corrected_proj_tiff)
        proj = tifffile.imread(chess_proj)
    else:
        print('proj array is not None, use it as is')
    
    if save_find_center is True:
        center_check_path = outpath +'/' + samplename + '/center_check/'
        print('saving find center value into ' +  center_check_path)   

        theta = tomopy.angles(proj.shape[0])
        
        if autocheck is True:
            check_cen_range_step = [int(proj.shape[2]/2)-50, int(proj.shape[2])/2+50, 5]

        if auto_selecslice is True:        
            cen_slice = int(proj.shape[1]/2)
        
        tomopy.write_center(proj, theta, 
                            center_check_path, cen_range = check_cen_range_step, ind = cen_slice, mask = True)
    
    return proj
#               
                                        
def srxfftomo_recon(corrected_proj_tiff = None, proj = None, 
                    rot_center = None, recon_algorithm = 'art', recon_section = None,
                    outpath = None, recon_outpath = 'recon', samplename = None):

    '''
    input needs to be either 
    1) corrected_proj_tiff, provided as a file path and name, 
    e.g. '/home/xf05id1/localdata/TomoCommissioning/testsample_01/testsample_01_corrected.tiff'
    it is a stack of tiff generated by srxfftomo_correction, where background and stage round out have been corrected
    2) projection numpy array, returned from srxfftomo_correction
    
    The 'proj' array assignment has priority. If it is None, proj was be read from corrected_proj_tiff 
    
    recon_section is a list of upper and lower bounds for y range, e.g. [380, 420]
    
    '''
    
    if proj is None:
        print('proj is not asigned, load corrected projection from:', chess_proj)
        proj = tifffile.imread(chess_proj)
    else:
        print('proj array is not None, use it as is')

    print('running reconstruction')

    if recon_section is not None:    
        proj = proj [:, recon_section[0]:recon_section[1], :]
    
    theta = tomopy.angles(proj.shape[0])
    rec = tomopy.recon(proj, theta, center=rot_center, algorithm= recon_algorithm)

    print('saving reconstruction into')
    recon_outfile = outpath +'/' + samplename + '/' + recon_outpath +'/' + samplename + '_recon'
    tomopy.write_tiff_stack(rec, fname=recon_outfile)
    
    return rec