forked from udacity/CarND-Vehicle-Detection
-
Notifications
You must be signed in to change notification settings - Fork 0
/
m4_test_video.py
46 lines (36 loc) · 1.23 KB
/
m4_test_video.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
#!/bin/sh
from moviepy.editor import VideoFileClip
import sys
import cv2
import vehicle_detection2
from extract_features import HogFeatureExtractor
import json
from sklearn.externals import joblib
class Env:
pass
class App:
def __init__(self, vehicle_detector):
self.w = 'preview'
cv2.namedWindow(self.w)
self.vehicle_detector = vehicle_detector
def process_image(self, img):
bgr = cv2.cvtColor(img, cv2.COLOR_RGB2BGR)
processed = self.vehicle_detector.process_image(bgr)
cv2.imshow(self.w, processed)
cv2.waitKey(33)
return cv2.cvtColor(processed, cv2.COLOR_BGR2RGB)
env_fname, clf_fname, input_fname, output_fname = sys.argv[1:]
with open(env_fname) as f:
env = json.load(f)
clf = joblib.load(clf_fname)
hog = HogFeatureExtractor(
cspace=env['colorspace'],
orient=env['orient'],
pix_per_cell=env['pix_per_cell'],
cell_per_block=env['cell_per_block'],
hog_channel=env['hog_channel'])
vehicle_detector = vehicle_detection2.VehicleDetector(clf, hog)
app = App(vehicle_detector)
input_clip = VideoFileClip(input_fname)
output_clip = input_clip.fl_image(app.process_image)
output_clip.write_videofile(output_fname, audio=False)