def main(image1_path, image2_path):
img1, gray1 = readImage(image1_path)
img2, gray2 = readImage(image2_path)
# load face model
conns = facetracker.LoadCon(r'external\FaceTracker\model\face.con')
trigs = facetracker.LoadTri(r'external\FaceTracker\model\face.tri')
tracker1 = facetracker.FaceTracker(
r'external\FaceTracker\model\face.tracker')
tracker2 = facetracker.FaceTracker(
r'external\FaceTracker\model\face.tracker')
tracker1.setWindowSizes((11, 9, 7))
tracker2.setWindowSizes((11, 9, 7))
# search feature points
feature_pair = getFeaturePoints(tracker1, tracker2, gray1, gray2)
intermediate_feature = interpolatePts(feature_pair)
frames = combineImages(intermediate_feature, TRIANGLES, image1_path,
image2_path)
frames.extend(frames[::-1])
while (True):
for i in range(0, len(frames)):
f = frames[i]
cv2.waitKey(20)
cv2.imshow("Cameras", f)
cv2.waitKey(20)
def get_tracker_points(args):
"""
FaceTracker settings:
- clamp : [0-4] 1 gives a very loose fit, 4 gives a very tight
fit
- iterations : Number of iterations. Should be a number in the
range: 1-25, 1 is fast and inaccurate, 25 is slow and
accurate
- tolerance : Matching tolerance. Should be a double in the range:
.01-1.
pyFaceTracker:
http://pythonhosted.org/pyFaceTracker/generated/facetracker.FaceTracker.html
FaceTracker:
https://github.com/kylemcdonald/FaceTracker
"""
portrait_path = args.portrait_path
tracker_path = args.tracker_path
clamp = args.clamp
iterations = args.iterations
tolerance = args.tolerance
out_dir = args.output_dir
plot = args.plot
# Load portrait
portrait = Image.open(portrait_path)
portrait_gray = portrait.convert("L")
portrait_gray_np = np.array(portrait_gray)
# Load tracker
tracker = facetracker.FaceTracker(tracker_path)
# Get tracker points
tracker.resetFrame()
if clamp:
tracker.clamp = clamp
tracker.iterations = iterations
if tolerance:
tracker.tolerance = tolerance
tracker.update(portrait_gray_np)
# Format tracker points
malformed_points = tracker.get2DShape()[0]
nb_coords = len(malformed_points)
nb_points = nb_coords / 2
xs = malformed_points[:nb_points]
ys = malformed_points[nb_points:]
points = zip(xs, ys)
points_np = np.squeeze(np.array(points))
# Save tracker points
name = rm_dir_and_ext(portrait_path)
np.save(out_dir + name + ".npy", points_np)
# Output plot
if plot:
save_plot(portrait, points_np, name, out_dir=out_dir)
def main(img_path):
#
# Load image
#
img = Image.open(img_path)
gray = img.convert('L')
img = np.asanyarray(img)
gray = np.asarray(gray)
#
# Load face model
#
conns = facetracker.LoadCon(r'..\external\FaceTracker\model\face.con')
trigs = facetracker.LoadTri(r'..\external\FaceTracker\model\face.tri')
tracker = facetracker.FaceTracker(
r'..\external\FaceTracker\model\face.tracker')
#
# Search for faces in the image
#
tracker.setWindowSizes((11, 9, 7))
if tracker.update(gray):
img = tracker.draw(img, conns, trigs)
obj3D = tracker.get3DShape()
fig3d = plt.figure()
ax = fig3d.add_subplot(111, projection='3d')
ax.scatter(obj3D[:66, 0], obj3D[66:132, 0], obj3D[132:, 0])
for i in range(66):
ax.text(obj3D[i], obj3D[i + 66], obj3D[i + 132], str(i))
ax.view_init(-90, -90)
else:
print 'Failed tracking face in image'
plt.figure()
plt.imshow(img)
plt.show()
def tutorial1():
img = Image.open(os.path.join(IMAGES_BASE, 'MonaLisa.jpg'))
gray = img.convert('L')
img = np.asanyarray(img)
gray = np.asarray(gray)
#
# Load face model
#
conns = facetracker.LoadCon(r'..\external\FaceTracker\model\face.con')
trigs = facetracker.LoadTri(r'..\external\FaceTracker\model\face.tri')
tracker = facetracker.FaceTracker(
r'..\external\FaceTracker\model\face.tracker')
#
# Search for faces in the image
#
tracker.setWindowSizes((11, 9, 7))
if tracker.update(gray):
img = tracker.draw(img, conns, trigs)
obj3D = tracker.get3DShape()
fig3d = plt.figure()
ax = fig3d.add_subplot(111, projection='3d')
ax.scatter(obj3D[:66, 0], obj3D[66:132, 0], obj3D[132:, 0])
for i in range(66):
ax.text(obj3D[i], obj3D[i + 66], obj3D[i + 132], str(i))
ax.view_init(-75, -90)
plt.title('3D Model of Mona Lisa\'s face')
savefig('3DMonaLisa')
else:
print 'Failed tracking face in image'
plt.figure()
plt.imshow(img)
plt.title('Face of the Mona Lisa')
savefig('MonaLisaFace')
# -*- encoding:utf-8 -*-
import os
import random
import string
import numpy as np
import cv2
import facetracker
from werkzeug.utils import secure_filename
from PIL import Image
conns = facetracker.LoadCon('./static/model/face.con')
trigs = facetracker.LoadTri('./static/model/face.tri')
tracker = facetracker.FaceTracker('./static/model/face.tracker')
cascade_path = "/usr/local/opt/opencv/share/OpenCV/haarcascades/haarcascade_frontalface_alt.xml"
allow_extends = set(['png', 'jpeg', 'jpg', 'gif'])
digit = 8
def allowed_file(filename):
return '.' in filename and \
filename.rsplit('.', 1)[1].lower() in allow_extends
def image_save(img):
if img and allowed_file(img.filename):
ImgName = img.filename
result = "ok"
img = cv2.imdecode(np.fromstring(img.read(), np.uint8),
pass
cv2.destroyAllWindows()
if __name__ == '__main__':
import sys, getopt
print help_message
args, video_src = getopt.getopt(sys.argv[1:], '', ['face=', 'con=', 'tri='])
try: video_src = video_src[0]
except: video_src = 0
args = dict(args)
face_fn = args.get('--con', r"../external/FaceTracker/model/face.tracker")
con_fn = args.get('--con', r"../external/FaceTracker/model/face.con")
tri_fn = args.get('--tri', r"../external/FaceTracker/model/face.tri")
tracker = facetracker.FaceTracker(face_fn)
conns = facetracker.LoadCon(con_fn)
trigs = facetracker.LoadTri(tri_fn)
cam = create_capture(video_src)
tracker.setWindowSizes((7,))
shape3D = np.random.randn(3, 66)
l = mlab.points3d(shape3D[0, :], shape3D[1, :], shape3D[2, :])
ms = l.mlab_source
print("Server is running on localhost:8001...")
server = pywsgi.WSGIServer(('0.0.0.0', 8001), myapp, handler_class=WebSocketHandler)
server.serve_forever()
from naoqi import ALProxy
import facetracker as track
tracker = track.FaceTracker()
tracker.startTracking()
#tts.say("")
def __init__(self, recipe):
self.recipe = recipe
self.faceTracker = tracker.FaceTracker(use_nao)
self.ledController = eye_leds.LedController(use_nao)