class ShotDetectorTestCase(unittest.TestCase):
def setUp(self):
logging.basicConfig(level=logging.ERROR)
video_uri = md().video_uri
print video_uri
self.video = Video(video_uri,begin=0,end=42)
self.sd = ShotDetector()
self.cuts,self.diss = self.sd.process(self.video.uri,profile='relax')
def test_cuts(self):
res = numpy.array([38,275,336,359,376,388,432,443,452,458,473,542,565,603])
assert_equal(self.cuts[:14],res)
def test_dissolves(self):
res = [[ 459, 465],
[ 713, 718],
[1293, 1301],
[1515, 1519],
[1605, 1613],
[1937, 1943],
[2184, 2192],
[2396, 2404],
[2439, 2444],
[2482, 2486],
[2488, 2494]]
assert_equal(self.diss,res)
def setUp(self):
logging.basicConfig(level=logging.ERROR)
video_uri = md().video_uri
print video_uri
self.video = Video(video_uri,begin=0,end=42)
self.sd = ShotDetector()
self.cuts,self.diss = self.sd.process(self.video.uri,profile='relax')
def getcuts(self):
u"""
video cut detection using GoodShotDetector class.
:param smoothbyhisto: if true, we remove cut if histogram difference isn't sufficient
:type: boolean
:rtype: list of frame number where cut was detected
"""
if 'goodshotdetection' not in self.hdf5['visual'].keys():
sd = ShotDetector()
cuts, diss = sd.process(self.uri)
cuts.sort()
diss.sort()
gsd = self.hdf5['visual'].create_group('goodshotdetection')
gsd.create_dataset('cuts', data=cuts)
gsd.create_dataset('dissolves', data=diss)
gsd = self.hdf5['visual']['goodshotdetection']
return map(numpy.array, (gsd['cuts'],gsd['dissolves']))
