lab 6

Author: paveldat

Methods for fractal analysis of digital images/lab 6/Readme.md

@@ -0,0 +1,19 @@
+# Task 6.
+Task: Calculate the value vector ln(𝐴𝛿)/ln(𝛿) 𝛿=1,..10. (You can use both symmetric and asymmetric methods).
+
+How to run:
+* Go to the folder containing the program code and image
+
+* Type in the console "python count_vector.py" or "python3 count_vector.py"
+
+* You will be prompted to enter the name of the image file located in the same folder as the program code. If you don't enter any name, the default program will try to find a file named "image.jpg"
+
+* To plot multiple files, enter multiple image names separated by spaces: "image1.jpg image2.jpg image3.jpg"
+
+* The image you submit for entry may be in color. The program itself will invert it into a black and white image, saving it with the name "inverted_" + the name of the original image
+
+* Next, the file "result_graph.png" will be created
+
+# Author
+Pavel Dat
+

Methods for fractal analysis of digital images/lab 6/count_vector.py

@@ -0,0 +1,96 @@
+from PIL import Image, ImageDraw
+import numpy as np
+import matplotlib.pyplot as plt
+
+delta_global =  np.arange(2, 11)
+CELL_SIZE = 20
+
+def convert_image(image, converted_file_name):
+    black_and_white = image.convert('1')
+    black_and_white.save(converted_file_name)
+    return black_and_white
+
+def get_A(cell, ds):
+    length, width = cell.size[0], cell.size[1]
+    u = np.zeros((len(ds) + 1, length + 1, width + 1))
+    b = np.zeros((len(ds) + 1, length + 1, width + 1))
+    v = np.zeros((len(ds) + 1))
+    A = np.zeros((len(ds) + 1))
+
+    for i in range(0, length):
+        for j in range(0, width):
+            u[0][i][j] = cell.getpixel((i, j))
+            b[0][i][j] = cell.getpixel((i, j))
+
+            for d in ds:
+                u[d][i][j] = max(u[d - 1][i][j] + 1, max(u[d - 1][i + 1][j + 1], u[d - 1][i - 1][j + 1], u[d - 1][i + 1][j - 1], u[d - 1][i - 1][j - 1]))
+                b[d][i][j] = min(b[d - 1][i][j] - 1, min(b[d - 1][i + 1][j + 1], b[d - 1][i - 1][j + 1], b[d - 1][i + 1][j - 1], b[d - 1][i - 1][j - 1]))
+    
+    for d in ds:
+        v[d] = 0
+        for i in range(0, length):
+            for j in range(0, width):
+                v[d] += u[d][i][j] - b[d][i][j]
+    
+    for d in ds:
+        A[d] = (v[d] - v[d-1]) / 2
+    
+    return A
+
+def count_dimension(image, delta):
+    length, width = image.size[0], image.size[1]
+    cell_length, cell_width = length//CELL_SIZE, width//CELL_SIZE
+    A = np.zeros((cell_length, cell_width, len(delta)+1))
+    for i in range(0, cell_length):
+        for j in range(0, cell_width):
+            area = (i*CELL_SIZE, j*CELL_SIZE, (i+1)*CELL_SIZE, (j+1)*CELL_SIZE)
+            cell = image.crop(area)
+            for d in delta:
+                A[i][j][d] = get_A(cell, delta)[d]
+
+    
+    As = []
+    for d in delta:
+        s = 0
+        for i in range(0, cell_length):
+            for j in range(0, cell_width):
+                s += A[i][j][d]
+        As.append(s)
+        
+    return np.polyfit(np.log(As), np.log(delta), 1)[0] * (-1)
+
+def get_graph_data(image):
+    data = []
+    for d in delta_global:
+        image_copy = image
+        data.append(count_dimension(image_copy, np.arange(1, d+1)))
+    return data
+
+
+
+if __name__ == '__main__':
+    finished = False
+    while not finished:
+        print("Please, type the name of the file. Or press enter to skip. Default name is image.jpg\n You can type multiple file names, separated by space\n Example: 'image1.jpg image2.jpg image3.jpg'")
+        names = input()
+        if not names:
+            names = ['image.jpg']
+        else:
+            names = names.split(' ')
+        try:
+            plt.cla()
+            fig, ax = plt.subplots()
+            for name in names:
+                im = Image.open(name)
+                im = convert_image(im, 'inverted_' + name)
+                file_data = get_graph_data(im)
+                ax.plot(delta_global, file_data, label = name)
+
+            ax.set_xlabel('delta')
+            ax.set_ylabel('ln(A delta)/ln(delta)')
+            ax.legend()
+            fig.savefig("result_graph.png")
+            finished = True
+        except Exception as e:
+            print(e)
+            print("Could not find the file. Please try again.")