car-recognition-program/imageRecognitionTrainingProgram.py at main · chasemalbright/car-recognition-program · 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
145
146
147
148
149
150
151
152
153
154
155
156
157
158
# python script that takes in database of images for traing and creates a model of what a vehicle is 'supposed' to look like
# time spent training is dependent upon how many image files are ran through the program.

# @chasealbright


from keras import applications
from keras import optimizers
from keras.models import Sequential, Model
from keras.layers import Dropout, Flatten, Dense, GlobalAveragePooling2D, Convolution2D, MaxPooling2D, Activation
from keras.optimizers import RMSprop, Adam, Adadelta
from keras.models import load_model
from keras import backend as k
from keras.callbacks import ModelCheckpoint, LearningRateScheduler, TensorBoard, EarlyStopping
import keras
from urllib.request import urlopen
import numpy as np
import cv2
from sklearn.model_selection import train_test_split
import pandas as pd

inputImageSize = [1280, 720]

def create_model():

    nb_conv = 5
    nb_filters = 8
    model = Sequential()
    model.add(Convolution2D(nb_filters, (nb_conv, nb_conv),
                            border_mode='valid',
                            input_shape=image_size, dim_ordering="tf") )
    model.add(Activation('relu'))

    model.add(MaxPooling2D(pool_size=(2,2)))

    model.add(Convolution2D(nb_filters, (nb_conv, nb_conv)))
    model.add(Activation('relu'))

    model.add(Convolution2D(nb_filters, (nb_conv, nb_conv)))
    model.add(Activation('relu'))
    model.add(MaxPooling2D(pool_size=(2,2)))

    model.add(Convolution2D(nb_filters, (nb_conv, nb_conv)))
    model.add(Activation('relu'))
    model.add(Dropout(0.25))

    model.add(Convolution2D(nb_filters*2, (nb_conv, nb_conv)))
    model.add(Activation('relu'))
    model.add(MaxPooling2D(pool_size=(2,2)))

    model.add(Convolution2D(nb_filters*2, (nb_conv, nb_conv)))
    model.add(Activation('relu'))

    model.add(Convolution2D(nb_filters*2, (nb_conv, nb_conv)))
    model.add(Activation('relu'))
    model.add(MaxPooling2D(pool_size=(2,2)))
    model.add(Convolution2D(nb_filters*2, (nb_conv, nb_conv)))
    model.add(Activation('relu'))
    model.add(Dropout(0.5))

    model.add(Flatten())
    model.add(Dense(256))
    model.add(Activation('relu'))
    model.add(Dropout(0.5))

    model.add(Dense(128))
    model.add(Activation('relu'))
    model.add(Dropout(0.5))

    model.add(Dense(1))
    model.add(Activation('linear'))

    model.compile(loss='mean_squared_error', optimizer=Adam())
    return model

def train_model(batch_size = 50, n_epoch = 20):
    train_data, train_target = read_and_normalize_train_data()

    cv_size = int(train_target.size/4)

    train_data[:,:,:] = train_data[0:,:,:]
    train_target = train_target[0:]

    X_train, X_valid, y_train, y_valid = train_test_split(train_data, train_target, test_size=cv_size, random_state=56741)

    model = create_model()
    model.fit(X_train, y_train, batch_size=batch_size, epochs=n_epoch, verbose=1, validation_data=(X_valid, y_valid) )

    predictions_valid = model.predict(X_valid, batch_size=50, verbose=1)
    compare = pd.DataFrame(data={'original':y_valid.reshape((cv_size,)),
             'prediction':predictions_valid.reshape((cv_size,))})
    compare.to_csv('compare.csv')

    return model

def load_train():
    x_train = []
    y_train = []

    datas = loadTrainingData(foldername)
    print('Read training images')
    i=0
    for row in datas:
        image_path = row[0]
        try:
            resp = urlopen(image_path)
            img = np.asarray(bytearray(resp.read()), dtype="uint8")
            img = cv2.cvtColor(cv2.imdecode(img, cv2.IMREAD_COLOR), cv2.COLOR_BGR2GRAY)
            img = cv2.resize(img, (image_size_c, image_size_r))
            x_train.append(img)
            y_train.append(int(row[1]))
            i=i+1
            if i>6000:
                break
        except:
            print(image_path)


    print('Training images read')
    return (np.expand_dims(x_train, axis=3), y_train)

def loadTrainingData(foldername):

    fullName = foldername + r'/freemontCorrectedResults.csv'
    file = open(fullName)
    datas = []
    for line in file:
        try:
            data = line.split(",")
            datas.append(data)
        except:
            print("")
    numel = len(datas)
    return datas

def read_and_normalize_train_data():
    train_data, train_target = load_train()
    train_data = np.array(train_data, dtype=np.float32)
    train_target = np.array(train_target, dtype=np.float32)
    print('Train shape:', train_data.shape)
    print(train_data.shape[0], 'train samples')
    return train_data, train_target

keras.backend.clear_session()
image_width = inputImageSize[0]
image_height = inputImageSize[1]
keras.backend.clear_session()
imageDownsizeFactor = 2
image_size_r = int(image_height/2)
image_size_c = int(image_width/2)

image_size = (image_size_r, image_size_c, 1)
foldername = r'//share/FreemontTraining'
model = train_model(n_epoch = 60)
model.save('FreemontModelsavetest.h5')

model = load_model('FreemontModelsavetest.h5')