-
Notifications
You must be signed in to change notification settings - Fork 2
/
Copy pathDrone_Net_Test.py
186 lines (159 loc) · 7.97 KB
/
Drone_Net_Test.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
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
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
import os
import tarfile
import threading
import cv2
import numpy as np
import six.moves.urllib as urllib
import tensorflow as tf
from object_detection.utils import label_map_util
from object_detection.utils import visualization_utils as vis_util
class Drone_Net_Test(threading.Thread):
def __init__(self, vision, process):
self.storage = []
self.last_box = None
self.process = process
self.img = None
self.boxes = None
self.coordinates = None
threading.Thread.__init__(self)
# Define the video stream
self.cap = cv2.VideoCapture(0) # Change only if you have more than one webcams
self.vision = vision
# What model to download.
# Models can bee found here:
# https://github.com/tensorflow/models/blob/master/research/object_detection/g3doc/detection_model_zoo.md
# MODEL_NAME = 'ssd_inception_v2_coco_2018_01_28'
MODEL_NAME = 'ssdlite_mobilenet_v2_coco_2018_05_09'
MODEL_FILE = MODEL_NAME + '.tar.gz'
DOWNLOAD_BASE = 'http://download.tensorflow.org/models/object_detection/'
# Path to frozen detection graph. This is the actual model that is used for the object detection.
PATH_TO_CKPT = MODEL_NAME + '/frozen_inference_graph.pb'
# List of the strings that is used to add correct label for each box.
PATH_TO_LABELS = os.path.join('object_detection/data', 'mscoco_label_map.pbtxt')
# Number of classes to detect
NUM_CLASSES = 90
# Download Model
opener = urllib.request.URLopener()
already_downloaded = input("Enter 'Y' if the network is already downloaded; 'N' if it isn't: ")
if (already_downloaded == "N"):
opener.retrieve(DOWNLOAD_BASE + MODEL_FILE, MODEL_FILE)
tar_file = tarfile.open(MODEL_FILE)
for file in tar_file.getmembers():
file_name = os.path.basename(file.name)
if 'frozen_inference_graph.pb' in file_name:
tar_file.extract(file, os.getcwd())
# Load a (frozen) Tensorflow model into memory.
self.detection_graph = tf.Graph()
with self.detection_graph.as_default():
od_graph_def = tf.GraphDef()
with tf.gfile.GFile(PATH_TO_CKPT, 'rb') as fid:
serialized_graph = fid.read()
od_graph_def.ParseFromString(serialized_graph)
tf.import_graph_def(od_graph_def, name='')
# Loading label map
# Label maps map indices to category names, so that when our convolution network predicts `5`,
# we know that this corresponds to `airplane`. Here we use internal utility functions,
# but anything that returns a dictionary mapping integers to appropriate string labels would be fine
label_map = label_map_util.load_labelmap(PATH_TO_LABELS)
categories = label_map_util.convert_label_map_to_categories(
label_map, max_num_classes=NUM_CLASSES, use_display_name=True)
self.category_index = label_map_util.create_category_index(categories)
# Helper code
def load_image_into_numpy_array(image):
(im_width, im_height) = image.size
return np.array(image.getdata()).reshape(
(im_height, im_width, 3)).astype(np.uint8)
# runs the neural net detection algorithm as a thread
def run(self):
config = tf.ConfigProto()
config.gpu_options.allow_growth = True
with self.detection_graph.as_default():
with tf.Session(graph=self.detection_graph, config=config) as sess:
while True:
# Read frame from camera
# print(image_np)
ret, image_np = self.cap.read()
if (image_np is not None):
# Expand dimensions since the model expects images to have shape: [1, None, None, 3]
image_np_expanded = np.expand_dims(image_np, axis=0)
# Extract image tensor
image_tensor = self.detection_graph.get_tensor_by_name('image_tensor:0')
# Extract detection boxes
boxes = self.detection_graph.get_tensor_by_name('detection_boxes:0')
# Extract detection scores
scores = self.detection_graph.get_tensor_by_name('detection_scores:0')
# Extract detection classes
classes = self.detection_graph.get_tensor_by_name('detection_classes:0')
# Extract number of detectionsd
num_detections = self.detection_graph.get_tensor_by_name(
'num_detections:0')
# Actual detection.
(boxes, scores, classes, num_detections) = sess.run(
[boxes, scores, classes, num_detections],
feed_dict={image_tensor: image_np_expanded})
# returns all box coordinates as a double array ([[ymin, xmin, ymax, xmax]]_
self.boxes = vis_util.get_box_coordinates(image_np,
np.squeeze(boxes),
np.squeeze(classes).astype(np.int32),
np.squeeze(scores),
self.category_index,
use_normalized_coordinates=True,
line_thickness=8,
only_get=1)
box = self.compute_boxes(self.boxes)
self.img = cv2.resize(image_np, (800, 600))
if (box is not None):
self.process.compute(box[0], box[1], box[2], box[3])
else:
self.process.set_pitch(0)
self.process.set_rotation(0)
self.process.set_tilt(0)
def compute_boxes(self, coordinates):
coordinates = np.array(coordinates)
if (not coordinates.sum() == 0):
coordinates = self.select_box(coordinates)
self.storage.append(coordinates)
if ((len(self.storage) > 30 or coordinates.sum() == 0) and len(self.storage) > 0):
self.storage.pop(0)
coordinates = np.array(self.storage, ndmin=2)
final_coordinates = []
if (coordinates.sum() == 0):
return None
else:
for i in range(len(coordinates[0])):
mean = 0
values = 0
for j in range(len(coordinates)):
mean += coordinates[j][i] * (j ** 5 + 1) # (j+1) weighs the input so last frame is worth more
values += (j ** 5 + 1)
final_coordinates.append(mean / values)
final_coordinates = np.array(final_coordinates)
self.last_box = final_coordinates
return final_coordinates
def select_box(self, boxes):
if (self.last_box is None):
self.last_box = boxes[0] # select box with highest certainty
else:
min = 99999999
min_index = 0
for i in range(len(boxes)):
if (abs(((boxes[i] - self.last_box) ** 2).sum()) < min):
min = abs(((boxes[i] - self.last_box) ** 2).sum())
min_index = i
#print("Tracking box: ", i, "difference:", min)
self.last_box = boxes[min_index]
return self.last_box
def get_boxes(self):
if self.boxes is None:
return None
else:
return np.array(self.boxes)
def get_adjusted_box(self):
if(len(self.storage)== 0):
return None
else:
return self.last_box
def get_image(self):
return self.img
# drone = Drone_Net()
# drone.start()