LandCoverClassification/main.py at master · echerif18/LandCoverClassification · GitHub

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
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
"""
Created on Thu Nov 26 15:12:46 2020

@author: eych
"""

import os
import sys

from sys import platform

from Models_1B import *
from Build_fit import *

from Models_2B import *
from Buid_2B import *


from LoadData import LoadData1B
from LoadData import LoadData2B

parent_path= sys.path.append(os.path.dirname(__file__))
print('Executable env: '+ str(sys.executable))


# Press the green button in the gutter to run the script.
if __name__ == '__main__':
    data_path = "D:/Tiles/Final_datasets/Sentinel_2_tiles_test/*/"
    print(data_path)
    print(platform)

    ###################Experiments with 1Branch-Models###############
    ###Data Loading

    ld=LoadData1B()
    ld.load_data(data_path)

    #rasterio
    training_labels=ld.labels_forTr(10)
    tr_set = ld.training_set10_Tr(10)

    ###Arcpy
    #training_labels=ld.labels_forTr_Arcpy(10)
    #tr_set=ld.training_set_arcpy(10)

    print(training_labels.shape)
    print(tr_set.shape)

    ####Parameters def
    nb_labels = training_labels.shape[3]
    print(nb_labels)
    # The dimensions of the input images
    nb_rows = training_labels.shape[1]
    print(nb_rows)
    nb_cols = training_labels.shape[2]
    print(nb_cols)

    ###class Weights definition for the weighted loss function
    ws = ld.weights_prep(training_labels, nb_rows, nb_cols)
    print(ws)


    training_images_nor = ld.norm(tr_set)
    print(len(training_images_nor))
    del tr_set

    training_images_nor_aug, training_labels_aug = ld.rot(training_images_nor, training_labels, 0, 5)
    del training_labels
    training_images_nor_aug, training_labels_aug = ld.flip(training_images_nor_aug, training_labels_aug, 5, 10)

    ####Parameters def
    channels = training_images_nor_aug.shape[3]
    print(channels)

    input_shape = (nb_rows, nb_cols, channels)


    #### Model building + Training
    model=DeepForestM2(input_shape, nb_labels)
    model.summary()
    name='DeepForestM2'
    loss='categorical_crossentropy'
    m=Buid_1B(nb_labels,model)
    print(m.nb_labels)
    m.build()
    #m.train_model(training_images_nor_aug,training_labels_aug,ws,name,loss,epochs=1)


    ###################Experiments with 2Branch-Models###############
    '''ld=LoadData2B()


    ld.load_data(data_path)
    training_labels=ld.labels_forTr(10)
    #training_labels=ld.labels_forTr_Arcpy(10)
    print(training_labels.shape)

    nb_labels = training_labels.shape[3]
    print(nb_labels)
    # The dimensions of the input images
    nb_rows = training_labels.shape[1]
    print(nb_rows)
    nb_cols = training_labels.shape[2]
    print(nb_cols)

    training_images_s1=ld.training_s1(10)
    training_images_s2=ld.training_s2(10)

    timestamps = training_images_s1.shape[1]
    print(timestamps)

    channels1= training_images_s1.shape[4]
    print(channels1)
    channels2= training_images_s2.shape[3]
    print(channels2)

    input_shape1=(timestamps,nb_rows,nb_cols, channels1)
    print(input_shape1)
    input_shape2= (nb_rows,nb_cols,channels2)
    print(input_shape2)

    ws=ld.weights_prep(training_labels,nb_rows,nb_cols)
    print(ws)


    training_images_s1_nor_aug,training_images_s2_nor_aug,training_labels_aug=ld.rot(training_images_s1,training_images_s2,training_labels,0,5)
    del training_labels
    training_images_s1_nor_aug,training_images_s2_nor_aug,training_labels_aug=ld.flip(training_images_s1,training_images_s2,training_labels_aug,5,10)

    # Model building

    model_glob, model1, model2 = two_branches(input_shape1, input_shape2, nb_labels)
    model_glob.summary()
    name = 'two_branch M'
    loss = 'categorical_crossentropy'
    m = Buid_2B(nb_labels, model_glob)
    print(m.nb_labels)
    m.build()'''
    #m.train_model(training_images_s1, training_images_s2, training_labels, weights=weights, valstart=15, name=name,loss=loss, epochs=1, batch_size=2)

    print('End Main')

# See PyCharm help at https://www.jetbrains.com/help/pycharm/