class TestHand():
def __init__(self):
self._kinect = Kinect()
self._body = BodyDetector()
self._hand = HandDetector(HandOtsu())
self._contour = HandContourDetector()
self._palm = PalmDetector()
def run(self):
for (depth, depth8, rgb) in self._kinect.get_data():
cv2.imshow('rgb', cv2.cvtColor(rgb, cv2.COLOR_BGR2RGB))
cv2.imshow('depth', depth8)
hand, mask = self._hand.run(depth, depth8)
(_, _, crop) = self._contour.run(mask)
if crop == None:
continue
cv2.imshow('hand', crop)
hand = self._palm.run(hand, crop)
if hand == None:
continue
cv2.imshow('hand final', hand)
cv2.waitKey(10)
def _crop(self, img, box):
crop = img[box[1]:box[1] + box[3], box[0]:box[0] + box[2]]
return crop
class TestPose():
def __init__(self, src):
self._src = src
self._kinect = Kinect()
self._body = BodyDetector()
self._hand = HandDetector()
self._contour = HandContourDetector()
self._pose = PoseClassifier(MultiPoseClassifier(src))
def run(self):
for (depth, depth8, rgb) in self._kinect.get_data():
contour = self._get_hand_contour(depth8, depth, rgb)
if contour.any():
self._contour.draw()
print self._pose.run(contour)
cv2.waitKey(5)
def _get_hand_contour(self, depth8, depth, rgb):
body = self._body.run(depth8)
(hand, _) = self._hand.run(body)
(cont, box, hc) = self._contour.run(hand)
if self._contour.not_valid():
return np.array([])
(cont, _, _) = self._contour.run(rgb, True, box, hc, depth)
return cont
class TestHand():
def __init__(self):
self._kinect = Kinect()
self._body = BodyDetector()
self._hand = HandDetector(HandOtsu())
self._contour = HandContourDetector()
self._palm = PalmDetector()
def run(self):
for (depth, depth8, rgb) in self._kinect.get_data():
cv2.imshow('rgb', cv2.cvtColor(rgb, cv2.COLOR_BGR2RGB))
cv2.imshow('depth', depth8)
hand, mask = self._hand.run(depth, depth8)
(_, _, crop) = self._contour.run(mask)
if crop == None:
continue
cv2.imshow('hand', crop)
hand = self._palm.run(hand, crop)
if hand == None:
continue
cv2.imshow('hand final', hand)
cv2.waitKey(10)
def _crop(self, img, box):
crop = img[box[1]:box[1]+box[3], box[0]:box[0]+box[2]]
return crop
class TestPose():
def __init__(self, src):
self._src = src
self._kinect = Kinect()
self._body = BodyDetector()
self._hand = HandDetector()
self._contour = HandContourDetector()
self._pose = PoseClassifier(MultiPoseClassifier(src))
def run(self):
for (depth, depth8, rgb) in self._kinect.get_data():
contour = self._get_hand_contour(depth8, depth, rgb)
if contour.any():
self._contour.draw()
print self._pose.run(contour)
cv2.waitKey(5)
def _get_hand_contour(self, depth8, depth, rgb):
body = self._body.run(depth8)
(hand, _) = self._hand.run(body)
(cont, box, hc) = self._contour.run(hand)
if self._contour.not_valid():
return np.array([])
(cont, _, _) = self._contour.run(rgb, True, box, hc, depth)
return cont
class TrainPose():
def __init__(self, id, nsamples, dst):
self._id = id
self._nsamples = nsamples
self._dst = dst
self._kinect = Kinect()
self._body = BodyDetector()
self._hand = HandDetector()
self._contour = HandContourDetector()
self._fdesc = FourierDescriptors()
self._train = []
def run(self):
warmup = True
for (depth8, depth, rgb) in self._kinect.get_data():
contour = self._get_hand_contour(depth8, depth, rgb)
if not contour:
continue
self._contour.draw()
if warmup:
key = cv2.waitKey(5)
if key == GO:
warmup = False
continue
fd = self._fdesc.run(contour)
self._train.append(fd)
if len(self._train) == self._nsamples:
self._save()
break
cv2.waitKey(5)
def _get_hand_contour(self, depth8, depth, rgb):
body = self._body.run(depth8)
(hand, _) = self._hand.run(body)
(cont, box, hc) = self._contour.run(hand)
if self._contour.not_valid():
return []
(cont, _, _) = self._contour.run(rgb, True, box, hc, depth)
return cont
def _save(self):
data = np.array(self._train)
model = EmpiricalCovariance().fit(np.array(self._train))
output = {'id': self._id, 'data': data, 'model': model}
pickle.dump(output, open(self._dst, 'wb'))
class TrainPose():
def __init__(self, id, nsamples, dst):
self._id = id
self._nsamples = nsamples
self._dst = dst
self._kinect = Kinect()
self._body = BodyDetector()
self._hand = HandDetector()
self._contour = HandContourDetector()
self._fdesc = FourierDescriptors()
self._train = []
def run(self):
warmup = True
for (depth8, depth, rgb) in self._kinect.get_data():
contour = self._get_hand_contour(depth8, depth, rgb)
if not contour:
continue
self._contour.draw()
if warmup:
key = cv2.waitKey(5)
if key == GO:
warmup = False
continue
fd = self._fdesc.run(contour)
self._train.append(fd)
if len(self._train) == self._nsamples:
self._save()
break
cv2.waitKey(5)
def _get_hand_contour(self, depth8, depth, rgb):
body = self._body.run(depth8)
(hand, _) = self._hand.run(body)
(cont, box, hc) = self._contour.run(hand)
if self._contour.not_valid():
return []
(cont, _, _) = self._contour.run(rgb, True, box, hc, depth)
return cont
def _save(self):
data = np.array(self._train)
model = EmpiricalCovariance().fit(np.array(self._train))
output = {'id': self._id, 'data': data, 'model': model}
pickle.dump(output, open(self._dst, 'wb'))
class TrainPose():
def __init__(self, id, nsamples, dst):
self._id = id
self._nsamples = nsamples
self._dst = dst
self._kinect = Kinect()
self._body = BodyDetector()
self._hand = HandDetector()
self._contour = HandContourDetector()
self._feature = GaborDescriptors(4, 4)
def run(self):
warmup = True
train = []
model = pickle.load(open('svm.pck', 'rb'))
for (depth, depth8, rgb) in self._kinect.get_data():
body = self._body.run(depth8)
(hand, _) = self._hand.run(body)
(cont, box, crop) = self._contour.run(hand)
hand = hand[box[1]:box[1]+box[3], box[0]:box[0]+box[2]]
cv2.imshow('hand', hand)
key = cv2.waitKey(2)
#if warmup:
# if key == GO:
# warmup = False
# continue
#if key != 97:
# continue
feature = self._feature.run(hand)
print model.predict(feature)
#train.append(feature)
#if len(train) == self._nsamples:
# self._save(train)
# break
#cv2.waitKey(2)
def _save(self, train):
data = np.array(train)
labels = self._id * np.ones(len(train), np.int)
output = {'labels': labels, 'data': data}
pickle.dump(output, open(self._dst, 'wb'))
i += 1
return (self._interval[0], self._interval[0] + i)
def _normalize(self, x):
xn = x / x.sum()
return xn
def show(self):
if self._body.any() and self._depth.any():
img = np.hstack((self._depth, self._body))
str = 'th1=%d th2=%d' % (self._interval[0], self._interval[1])
cv2.imshow(str, img)
if __name__ == '__main__':
from kinect import Kinect
kinect = Kinect()
body = BodyDetector()
for (d, _) in kinect.get_data():
b = body.run(d)
#body.show()
if kinect.stop(kinect.wait()):
break
#import sys; sys.stdin.read(1)
return (self._interval[0], self._interval[0] + i)
def _normalize(self, x):
xn = x / x.sum()
return xn
def show(self):
if self._body.any() and self._depth.any():
img = np.hstack((self._depth, self._body))
str = 'th1=%d th2=%d' %(self._interval[0], self._interval[1])
cv2.imshow(str, img)
if __name__ == '__main__':
from kinect import Kinect
kinect = Kinect()
body = BodyDetector()
for (d, _) in kinect.get_data():
b = body.run(d)
#body.show()
if kinect.stop(kinect.wait()):
break
#import sys; sys.stdin.read(1)
from kinect import Kinect
from body import BodyDetector
from hand import HandDetector, HandContourDetector
from pose import PoseClassifier, OpenCloseClassifier, MultiPoseClassifier
kinect = Kinect()
body = BodyDetector()
hand = HandDetector()
contour = HandContourDetector()
#pose = PoseClassifier(OpenCloseClassifier())
pose = PoseClassifier(MultiPoseClassifier())
for (depth, depth8, rgb) in kinect.get_data():
b = body.run(depth8)
(h, _) = hand.run(b)
#cv2.imshow('hand', h)
(ccc, box, hc) = contour.run(h)
if len(ccc) < 100:
continue
(ccc, _, _) = contour.run(rgb, True, box, hc, depth)
p = pose.run(ccc)
if p == -1:
from kinect import Kinect
from body import BodyDetector
from hand import HandDetector, HandContourDetector
from pose import PoseClassifier, OpenCloseClassifier, MultiPoseClassifier
kinect = Kinect()
body = BodyDetector()
hand = HandDetector()
contour = HandContourDetector()
# pose = PoseClassifier(OpenCloseClassifier())
pose = PoseClassifier(MultiPoseClassifier())
for (depth, depth8, rgb) in kinect.get_data():
b = body.run(depth8)
(h, _) = hand.run(b)
# cv2.imshow('hand', h)
(ccc, box, hc) = contour.run(h)
if len(ccc) < 100:
continue
(ccc, _, _) = contour.run(rgb, True, box, hc, depth)
p = pose.run(ccc)
if p == -1: