-
Notifications
You must be signed in to change notification settings - Fork 0
/
process_data.py
75 lines (59 loc) · 2.4 KB
/
process_data.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
import os
import shutil
import csv
import random
from PIL import Image, ImageOps
IN_DIR = "D:\Documents\TrackmaniaSelfDrivingData"
OUT_DIR = "data"
IN_DATA_FILE = "telemetry.csv"
OUT_DATA_FILE = "labels.csv"
OUT_TRAIN_FILE = "train.csv"
OUT_TEST_FILE = "test.csv"
TRAIN_TEST_SPLIT = 0.85
RESIZE = (64, 64)
def write_data_row(writer, img, file_name, speed, steering):
writer.writerow({"img_file": file_name, "speed": speed, "steering": steering})
img.save(os.path.join(OUT_DIR, file_name))
def convert_raw_to_output():
if os.path.isdir(OUT_DIR):
shutil.rmtree(OUT_DIR)
os.mkdir(OUT_DIR)
with open(os.path.join(IN_DIR, IN_DATA_FILE), newline="") as csv_in_file, open(
os.path.join(OUT_DIR, OUT_DATA_FILE), "w", newline=""
) as csv_out_file:
reader = csv.DictReader(csv_in_file)
fieldnames = ["img_file", "speed", "steering"]
writer = csv.DictWriter(csv_out_file, fieldnames=fieldnames)
writer.writeheader()
rows = list(reader)
for i, row in enumerate(rows):
img = Image.open(os.path.join(IN_DIR, row["img_file"]))
if RESIZE:
img = img.resize(RESIZE)
write_data_row(writer, img, row["img_file"], row["speed"], row["steering"])
write_data_row(
writer,
ImageOps.mirror(img),
os.path.splitext(row["img_file"])[0] + "_flip.png",
row["speed"],
-float(row["steering"]),
)
if i % 100 == 0:
print(str((i + 1) / len(rows) * 100) + '%')
def split_data():
with open(os.path.join(OUT_DIR, OUT_DATA_FILE), newline="") as csv_label_file, \
open(os.path.join(OUT_DIR, OUT_TRAIN_FILE), 'w', newline="") as csv_train_file, \
open(os.path.join(OUT_DIR, OUT_TEST_FILE), 'w', newline="") as csv_test_file:
reader = csv.DictReader(csv_label_file)
train_writer = csv.DictWriter(csv_train_file, fieldnames=reader.fieldnames)
test_writer = csv.DictWriter(csv_test_file, fieldnames=reader.fieldnames)
train_writer.writeheader()
test_writer.writeheader()
rows = list(reader)
random.shuffle(rows)
train = rows[:int((len(rows)) * TRAIN_TEST_SPLIT)]
test = rows[int((len(rows)) * TRAIN_TEST_SPLIT):]
train_writer.writerows(train)
test_writer.writerows(test)
convert_raw_to_output()
split_data()