1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
|
"""!
@brief Examples of usage and demonstration of abilities of Phase Oscillatory Neural Network based on Kuramoto model.
@authors Andrei Novikov (pyclustering@yandex.ru)
@date 2014-2020
@copyright BSD-3-Clause
"""
from pyclustering.samples.definitions import IMAGE_SYMBOL_SAMPLES;
from pyclustering.utils import read_image, rgb2gray;
from pyclustering.nnet import solve_type;
from pyclustering.nnet.syncpr import syncpr, syncpr_visualizer;
import random;
import math;
def template_recognition_image(images, steps, time, corruption = 0.1):
samples = [];
for file_name in images:
data = read_image(file_name);
image_pattern = rgb2gray(data);
for index_pixel in range(len(image_pattern)):
if (image_pattern[index_pixel] < 128):
image_pattern[index_pixel] = 1;
else:
image_pattern[index_pixel] = -1;
samples += [ image_pattern ];
net = syncpr(len(samples[0]), 0.3, 0.3, ccore = True);
net.train(samples);
# Recognize the each learned pattern
for i in range(len(samples)):
sync_output_dynamic = net.simulate(steps, time, samples[i], solve_type.RK4, True);
syncpr_visualizer.show_output_dynamic(sync_output_dynamic);
syncpr_visualizer.show_pattern(sync_output_dynamic, 10, 10);
# corrupt a little bit by black and white pixels
for _ in range( math.floor(len(samples[i]) * corruption) ):
random.seed();
random_pixel = math.floor(random.random() * len(samples[i]));
samples[i][random_pixel] = 1;
random_pixel = math.floor(random.random() * len(samples[i]));
samples[i][random_pixel] = -1;
sync_output_dynamic = net.simulate(steps, time, samples[i], solve_type.RK4, True);
syncpr_visualizer.show_output_dynamic(sync_output_dynamic);
syncpr_visualizer.show_pattern(sync_output_dynamic, 10, 10);
syncpr_visualizer.animate_pattern_recognition(sync_output_dynamic, 10, 10, title = "Pattern Recognition");
def small_mind_image_recognition():
"""!
@brief Trains network using letters 'M', 'I', 'N', 'D' and recognize each of them with and without noise.
"""
images = [];
images += IMAGE_SYMBOL_SAMPLES.LIST_IMAGES_SYMBOL_M;
images += IMAGE_SYMBOL_SAMPLES.LIST_IMAGES_SYMBOL_I;
images += IMAGE_SYMBOL_SAMPLES.LIST_IMAGES_SYMBOL_N;
images += IMAGE_SYMBOL_SAMPLES.LIST_IMAGES_SYMBOL_D;
template_recognition_image(images, 100, 10, 0.2);
def small_abc_image_recognition():
"""!
@brief Trains network using letters 'A', 'B', 'C', and recognize each of them with and without noise.
"""
images = [];
images += IMAGE_SYMBOL_SAMPLES.LIST_IMAGES_SYMBOL_A;
images += IMAGE_SYMBOL_SAMPLES.LIST_IMAGES_SYMBOL_B;
images += IMAGE_SYMBOL_SAMPLES.LIST_IMAGES_SYMBOL_C;
template_recognition_image(images, 250, 25);
def small_ftk_image_recognition():
"""!
@brief Trains network using letters 'F', 'T', 'K' and recognize each of them with and without noise.
"""
images = [];
images += IMAGE_SYMBOL_SAMPLES.LIST_IMAGES_SYMBOL_F;
images += IMAGE_SYMBOL_SAMPLES.LIST_IMAGES_SYMBOL_T;
images += IMAGE_SYMBOL_SAMPLES.LIST_IMAGES_SYMBOL_K;
template_recognition_image(images, 100, 10, 0.2);
small_mind_image_recognition();
small_abc_image_recognition();
small_ftk_image_recognition();
|
