src/graphical_analysis.py

   1 #coding: utf-8
   2 #file name: graphical_analysis_pyqt4.py
   3 #Laueye - A program to analyze Laue diffraction diagrams.
   4 #Copyright (C) <2008>  <Grissiom>
   5 #
   6 #This program is free software: you can redistribute it and/or modify
   7 #it under the terms of the GNU General Public License as published by
   8 #the Free Software Foundation, either version 3 of the License, or
   9 #(at your option) any later version.
  10 #
  11 #This program is distributed in the hope that it will be useful,
  12 #but WITHOUT ANY WARRANTY; without even the implied warranty of
  13 #MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  14 #GNU General Public License for more details.
  15 #
  16 #You should have received a copy of the GNU General Public License
  17 #along with this program.  If not, see <http://www.gnu.org/licenses/>.
  18
  19 import Image
  20 import ImageFilter
  21 import numpy
  22
  23 class circle:
  24         '''a circle's data'''
  25         def __init__(self):
  26                 self.edge = []
  27                 self.x_boundry = []
  28                 self.y_boundry = []
  29         def center(self):
  30                 '''return the location of the circle, in flaot format'''
  31                 x = 0
  32                 y = 0
  33                 for i in self.edge:
  34                         x += i[0]
  35                         y += i[1]
  36                 return float(x) / len(self.edge), float(y) / len(self.edge)
  37
  38 scope = 2
  39 '''设置程序扫描的范围'''
  40
  41 cl_size_min = 5
  42 '''define the min size of a circle, used to swapt put noise.'''
  43
  44 cl_size_max = 500
  45 '''define the max size of a circle, used to swapt put noise.'''
  46
  47 threshold = 7
  48 '''define the threshold of the spot'''
  49
  50 def figure_out_circles(im_in):
  51         '''figure out the circles in it
  52
  53         @param im_in the img to analyz
  54         @return return a list of circles
  55         '''
  56         im_in = denoise_img(im_in)
  57         gray_table = conv_to_array(im_in)
  58         width, height = im_in.size
  59         assert gray_table.shape == (height, width)
  60         status = []
  61         for i in range(height):
  62                 status.append([])
  63                 for k in range(width):
  64                         status[i].append(0)
  65         find_edge(gray_table, status, im_in)
  66         return _get_circles(status)
  67
  68 def denoise_img(im_in):
  69         '''return a denoised img
  70
  71         @param im_in img in
  72         @return im_out img out'''
  73         im_out = im_in.filter(ImageFilter.BLUR)
  74         im_out = im_out.filter(ImageFilter.BLUR)
  75         return im_out
  76
  77 def conv_to_array(im_in):
  78         '''convert the given img to numpy.narray
  79
  80         convert the given img to numpy.narray, each item indicates the gray scale.
  81         @param im_in the img to convert
  82         @return nar narray that convert from img'''
  83         width, height = im_in.size
  84         nar = numpy.zeros((height, width), 'l')
  85         im_in = im_in.convert('L')
  86         pix = im_in.load()
  87         for i in range(width):
  88                 for k in range(height):
  89                         nar[k, i] = pix[i, k]
  90         return nar
  91
  92 def find_edge(nar_in, status, im_in):
  93         '''determine edge in nar_in.
  94
  95         set edge status to 2.
  96         @param nar_in the gray scaled matrix
  97         @param status the status table to set up'''
  98         height, width = nar_in.shape
  99         size = 2
 100         k = size
 101         while k < width - size:
 102                 i = size
 103                 while i < height - size:
 104                         gr1 = numpy.average(nar_in[i - size:i, k - size:k])
 105                         gr2 = numpy.average(nar_in[i:i + size, k - size:k])
 106                         gr3 = numpy.average(nar_in[i:i + size, k:k + size])
 107                         gr4 = numpy.average(nar_in[i - size:i, k:k + size])
 108                         if max(gr1, gr2, gr3, gr4) - min(gr1,  gr2,  gr3,  gr4) > threshold:
 109                                 status[i][k] = 2
 110                                 im_in.putpixel((k, i), (255, 0, 0))
 111                         i += size
 112                 k += size
 113         im_in.save('../test/test-pred.png')
 114
 115
 116 def _get_boundry(in_list, status):
 117         '''get the boundry of the whole circle based on the status table.
 118
 119         @param in_list a point on the circle.
 120         @param status the status table which is the base to analyze
 121         @return edge a list of tuples which are the edge of the circle, in (x, y) format.
 122         '''
 123         edge = []
 124         while len(in_list) != 0:
 125                 y, x = in_list.pop()
 126                 edge.append((y, x))
 127                 status[y][x] = 4
 128                 for i in range(-scope, scope + 1):
 129                         for k in range(-scope, scope + 1):
 130                                 if status[y + i][x + k] == 2:
 131                                         in_list.append([y + i, x + k])
 132         return edge
 133
 134 def _get_circles(status):
 135         '''get the circles in the whole image.
 136
 137         @param status the status table which is the base of analyze.
 138         @return a list of circles.'''
 139         width = len(status)
 140         height = len(status[0])
 141         #TODO:maybe we need to check whether im_status is a matrices here.
 142         # i.e, all the len(im_status[:]) have the same value.
 143         circles = []
 144         for i in range(scope,  width - scope):
 145                 for k in range(scope,  height - scope):
 146                         if status[k][i] == 2:
 147                                 circles.append(circle())
 148                                 circles[-1].edge = _get_boundry([[k, i]], status)
 149                                 if len(circles[-1].edge) < cl_size_min or len(circles[-1].edge) > cl_size_max:
 150                                         del circles[-1]
 151         return circles
 152
 153 if __name__ == '__main__':
 154         circles = figure_out_circles(Image.open('../test/test_laue.jpg'))
 155 #       circles = figure_out_circles(Image.open('../test/test.png'))
 156         print 'we have', len(circles),'circles'
 157         #print 'thay are', [i.center() for i in circles]
 158 else:
 159         pass
 160 pass
 161