-
Notifications
You must be signed in to change notification settings - Fork 4
/
doublecheck.py
125 lines (105 loc) · 3.81 KB
/
doublecheck.py
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
# USAGE
# python test_network.py --model santa_not_santa.model --image images/examples/santa_01.png
# import the necessary packages
import os
os.environ['KERAS_BACKEND'] = 'theano'
import numpy as np
import argparse
import imutils
import cv2
import PIL.Image as Image
import PIL.ImageDraw as ImageDraw
from keras.preprocessing.image import img_to_array
from keras.models import load_model
import math
class DoubleChecker:
def __init__(self):
self.MODEL_NAME = "safe_unsafe.model"
self.model = load_model(self.MODEL_NAME)
def double_check(self, image, cropbox, use_normalized_coordinates=True):
# cv2.imshow("test", image)
# cv2.waitKey()
# print(cropbox)
cropbox = self.normalize_cropbox(image, cropbox, use_normalized_coordinates)
cropped_image = self.crop(image, cropbox)
# cv2.imshow("test", cropped_image)
# cv2.waitKey()
# print(cropbox)
# pre-process the image for classification
try:
cropped_image = cv2.resize(cropped_image, (28, 28))
# cv2.imshow("test", cropped_image)
# cv2.waitKey()
cropped_image = cropped_image.astype("float") / 255.0
cropped_image = img_to_array(cropped_image)
cropped_image = np.expand_dims(cropped_image, axis=0)
safe, unsafe = self.model.predict(cropped_image)[0]
label = "unsafe" if safe > unsafe else "safe"
proba = safe if safe > unsafe else unsafe
except:
label = "safe"
proba = 0.25
# print((label, proba))
# label = "{}: {:.2f}%".format(label, proba * 100)
return (label, proba)
# def normalize_cropbox(self, image, cropbox):
# draw = ImageDraw.Draw(image)
# im_width, im_height = image.size
# ymin, xmin, ymax, xmax = cropbox
# if use_normalized_coordinates:
# (left, right, top, bottom) = (xmin * im_width, xmax * im_width,
# ymin * im_height, ymax * im_height)
# else:
# (left, right, top, bottom) = (xmin, xmax, ymin, ymax)
# return (ymin, xmin, ymax, xmax)
def normalize_cropbox(self, image, cropbox, use_normalized_coordinates=True):
# draw = ImageDraw.Draw(image)
im_width, im_height = 1920, 1080
ymin, xmin, ymax, xmax = cropbox
(left, right, top, bottom) = (xmin * im_width, xmax * im_width,
ymin * im_height, ymax * im_height)
return (top, bottom, left, right)
def crop(self, numpy_image, crop_box):
# print(crop_box)
(top, bottom, left, right) = crop_box
top = math.floor(top)
bottom = math.floor(bottom)
left = math.floor(left)
right = math.floor(right)
# print("Cropbox ", (top, bottom, left, right))
# x1 = int(crop_box[0])-25
# y1 = int(crop_box[1])-25
# x2 = int(crop_box[2])+25
# y2 = int(crop_box[3])+25
return numpy_image[top:bottom, left:right]
# # construct the argument parse and parse the arguments
# ap = argparse.ArgumentParser()
# ap.add_argument("-m", "--model", required=True,
# help="path to trained model model")
# ap.add_argument("-i", "--image", required=True,
# help="path to input image")
# args = vars(ap.parse_args())
# # load the image
# image = cv2.imread(args["image"])
# orig = image.copy()
# # pre-process the image for classification
# image = cv2.resize(image, (28, 28))
# image = image.astype("float") / 255.0
# image = img_to_array(image)
# image = np.expand_dims(image, axis=0)
# # load the trained convolutional neural network
# print("[INFO] loading network...")
# model = load_model(args["model"])
# # classify the input image
# (notSanta, santa) = model.predict(image)[0]
# # build the label
# label = "Safe" if santa > notSanta else "Not Safe"
# proba = santa if santa > notSanta else notSanta
# label = "{}: {:.2f}%".format(label, proba * 100)
# # draw the label on the image
# output = imutils.resize(orig, width=400)
# cv2.putText(output, label, (10, 25), cv2.FONT_HERSHEY_SIMPLEX,
# 0.7, (0, 255, 0), 2)
# # show the output image
# cv2.imshow("Output", output)
# cv2.waitKey(0)