-
Notifications
You must be signed in to change notification settings - Fork 14
/
get_segmentation.py
151 lines (128 loc) · 8.41 KB
/
get_segmentation.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
from __future__ import print_function, absolute_import, division
import numpy as np
import tensorflow as tf
import tensorflow.contrib.eager as tfe
import os
import nets.Network as Segception
import argparse
from utils.utils import get_params, restore_state, init_model, inference
import cv2
from collections import namedtuple
# enable eager mode
tf.enable_eager_execution()
tf.set_random_seed(7)
np.random.seed(7)
Label = namedtuple( 'Label' , [
'name' , # The identifier of this label, e.g. 'car', 'person', ... .
# We use them to uniquely name a class
'eventId' , # An integer ID that is associated with this label.
# The IDs are used to represent the label in ground truth images
# An ID of -1 means that this label does not have an ID and thus
# is ignored when creating ground truth images (e.g. license plate).
# Do not modify these IDs, since exactly these IDs are expected by the
# evaluation server.
'trainId' , # Feel free to modify these IDs as suitable for your method. Then create
# ground truth images with train IDs, using the tools provided in the
# 'preparation' folder. However, make sure to validate or submit results
# to our evaluation server using the regular IDs above!
# For trainIds, multiple labels might have the same ID. Then, these labels
# are mapped to the same class in the ground truth images. For the inverse
# mapping, we use the label that is defined first in the list below.
# For example, mapping all void-type classes to the same ID in training,
# might make sense for some approaches.
# Max value is 255!
'category' , # The name of the category that this label belongs to
'categoryId' , # The ID of this category. Used to create ground truth images
# on category level.
'color' , # The color of this label
] )
labels = [
# name eventId trainId category catId hasInstances ignoreInEval color
Label( 'unlabeled' , 255 , 255 , 'void' , 0 , ( 0, 0, 0) ),
Label( 'ego vehicle' , 255 , 255 , 'void' , 0 , ( 0, 0, 0) ),
Label( 'rectification border' , 255 , 255 , 'void' , 0 , ( 0, 0, 0) ),
Label( 'out of roi' , 255 , 255 , 'void' , 0 , ( 0, 0, 0) ),
Label( 'static' , 255 , 255 , 'void' , 0 , ( 0, 0, 0) ),
Label( 'dynamic' , 255 , 255 , 'void' , 0 , (111, 74, 0) ),
Label( 'ground' , 255 , 255 , 'void' , 0 , ( 81, 0, 81) ),
Label( 'road' , 0 , 0 , 'flat' , 1 , (128, 64,128) ),
Label( 'sidewalk' , 0 , 1 , 'flat' , 1 , (244, 35,232) ),
Label( 'parking' , 0 , 255 , 'flat' , 1 , (250,170,160) ),
Label( 'rail track' , 0 , 255 , 'flat' , 1 , (230,150,140) ),
Label( 'building' , 1 , 2 , 'construction' , 2 , ( 70, 70, 70) ),
Label( 'wall' , 1 , 3 , 'construction' , 2 , (102,102,156) ),
Label( 'fence' , 1 , 4 , 'construction' , 2 , (190,153,153) ),
Label( 'guard rail' , 1 , 255 , 'construction' , 2 , (180,165,180) ),
Label( 'bridge' , 1 , 255 , 'construction' , 2 , (150,100,100) ),
Label( 'tunnel' , 1 , 255 , 'construction' , 2 , (150,120, 90) ),
Label( 'pole' , 2 , 5 , 'object' , 3 , (153,153,153) ),
Label( 'polegroup' , 2 , 255 , 'object' , 3 , (153,153,153) ),
Label( 'traffic light' , 2 , 6 , 'object' , 3 , (250,170, 30) ),
Label( 'traffic sign' , 2 , 7 , 'object' , 3 , (220,220, 0) ),
Label( 'vegetation' , 3 , 8 , 'nature' , 4 , (107,142, 35) ),
Label( 'terrain' , 3 , 9 , 'nature' , 4 , (152,251,152) ),
Label( 'sky' , 1 , 10 , 'sky' , 5 , ( 70,130,180) ),
Label( 'person' , 4 , 11 , 'human' , 6 , (220, 20, 60) ),
Label( 'rider' , 4 , 12 , 'human' , 6 , (255, 0, 0) ),
Label( 'car' , 5 , 13 , 'vehicle' , 7 , ( 0, 0,142) ),
Label( 'truck' , 5 , 14 , 'vehicle' , 7 , ( 0, 0, 70) ),
Label( 'bus' , 5 , 15 , 'vehicle' , 7 , ( 0, 60,100) ),
Label( 'caravan' , 5 , 255 , 'vehicle' , 7 , ( 0, 0, 90) ),
Label( 'trailer' , 5 , 255 , 'vehicle' , 7 , ( 0, 0,110) ),
Label( 'train' , 5 , 16 , 'vehicle' , 7 , ( 0, 80,100) ),
Label( 'motorcycle' , 5 , 17 , 'vehicle' , 7 , ( 0, 0,230) ),
Label( 'bicycle' , 5 , 18 , 'vehicle' , 7 , (119, 11, 32) ),
Label( 'license plate' , 255 , -1 , 'vehicle' , 7 , ( 0, 0,142) ),
]
trainId2label = { label.trainId : label for label in reversed(labels) }
def fromIdTrainToId(imgin):
imgout = imgin.copy()
for idTrain in trainId2label:
imgout[imgin == idTrain] = trainId2label[idTrain].eventId
return imgout
if __name__ == "__main__":
parser = argparse.ArgumentParser()
parser.add_argument("--image_path", help="image path", default='/media/snowflake/Data/city/images/train/aachen_000003_000019_gtFine_labelIds.png')
parser.add_argument("--model_path", help="Model path", default='weights/cityscapes_grayscale')
parser.add_argument("--n_classes", help="number of classes to classify", default=19)
parser.add_argument("--width", help="number of epochs to train", default=352)
parser.add_argument("--height", help="number of epochs to train", default=224)
parser.add_argument("--n_gpu", help="number of the gpu", default=0)
args = parser.parse_args()
# some parameters
os.environ["CUDA_VISIBLE_DEVICES"] = str(args.n_gpu)
n_classes = int(args.n_classes)
width = int(args.width)
height = int(args.height)
channels = 1
name_best_model = os.path.join(args.model_path, 'best')
# build model and optimizer
model = Segception.Segception_v4(num_classes=n_classes, weights=None, input_shape=(None, None, channels))
# Init models (optional, just for get_params function)
init_model(model, input_shape=(1, width, height, channels))
variables_to_restore = model.variables
variables_to_save = model.variables
variables_to_optimize = model.variables
# Init saver. can use also ckpt = tfe.Checkpoint((model=model, optimizer=optimizer,learning_rate=learning_rate, global_step=global_step)
saver_model = tfe.Saver(var_list=variables_to_save)
restore_model = tfe.Saver(var_list=variables_to_restore)
# restore if model saved and show number of params
restore_state(restore_model, name_best_model)
get_params(model)
img = cv2.imread(args.image_path, 0)
img = cv2.resize(img, (width, height), interpolation=cv2.INTER_AREA).astype(np.float32)
img = np.expand_dims(img, -1)
img = np.expand_dims(img, 0)
print(img.shape)
prediction = inference(model, img, n_classes, flip_inference=True, scales=[0.75, 1, 1.5], preprocess_mode=None)
print(prediction.numpy().shape)
prediction = tf.argmax(prediction, -1)
print(prediction.numpy().shape)
img = np.squeeze(img).astype(np.uint8)
prediction = np.squeeze(prediction.numpy()).astype(np.uint8)
prediction_6classes = fromIdTrainToId(prediction).astype(np.uint8)
cv2.imshow('image', img)
cv2.imshow('pred (cityscapes classes)', prediction*13) # *13 for visualization
cv2.imshow('pred (event classes)', prediction_6classes*40)# *40 for visualization
cv2.waitKey(0)
cv2.destroyAllWindows()