def get_faces(image_path, api_mode=False, rescale_face_crop=0):
"""
Return a list of cropped faces given an image path
:param image_path: Path to image
:type image_path: str
:param api_mode: If api_mode is True get_faces returns a list of found HaarFeatures
:type api_mode: bool
:returns: list of images
"""
original_image = Image(image_path)
faces = original_image.findHaarFeatures(segment_face)
if api_mode:
return faces
else:
if rescale_face_crop:
return [
original_image.crop(
scale_bounding_box(face.boundingBox(), rescale_face_crop))
for face in faces
]
else:
return [face.crop() for face in faces]
def line_blobs(id):
cur = g.db.execute(
'SELECT id, x, y, width, height FROM blobs WHERE image=?', [id])
blobs = cur.fetchall()
entries = []
img = Image("static\\img\\%d.jpg" % id)
for i, entry in enumerate(blobs):
blob = img.crop(entry[1], entry[2], entry[3], entry[4])
if blob and 'SK_MODEL' in app.config:
if blob.height > blob.width:
blob = blob.resize(h=app.config['PATCH_SIZE'])
else:
blob = blob.resize(w=app.config['PATCH_SIZE'])
blob = blob.embiggen(
(app.config['PATCH_SIZE'], app.config['PATCH_SIZE']))
np_img = blob.getGrayNumpy().transpose().reshape(-1)
pred = labels.inverse_transform(sk_model.predict(np_img))[0]
if app.config['DEBUG']:
blob.save("tmp\\pic%d %s.jpg" % (i, pred))
entries.append(
dict(x=entry[1],
y=entry[2],
width=entry[3],
height=entry[4],
pred=pred))
else:
entries.append(
dict(x=entry[1], y=entry[2], width=entry[3], height=entry[4]))
return jsonify(blobs=entries)
def capture(self):
count = 0
currentframes = []
self.framecount = self.framecount + 1
for c in self.cameras:
img = ""
if c.__class__.__name__ == "Kinect" and c._usedepth == 1:
img = c.getDepth()
elif c.__class__.__name__ == "Kinect" and c._usematrix == 1:
mat = c.getDepthMatrix().transpose()
img = Image(np.clip(mat - np.min(mat), 0, 255))
else:
img = c.getImage()
if self.config.cameras[0].has_key('crop'):
img = img.crop(*self.config.cameras[0]['crop'])
frame = M.Frame(capturetime = datetime.utcnow(),
camera= self.config.cameras[count]['name'])
frame.image = img
currentframes.append(frame)
while len(self.lastframes) > self.config.max_frames:
self.lastframes.pop(0)
# log.info('framecount is %s', len(self.lastframes))
# Session().redis.set("framecount", self.framecount)
count = count + 1
self.lastframes.append(currentframes)
return currentframes
def crop():
#Begin Processing image
print ('Begin processing image...')
fullImage = Image('test/image/expected/fullImage.jpg')
hundreds = fullImage.crop(602, 239, 28, 63)
hundreds = hundreds.binarize(65).invert()
hundreds.save('test/image/actual/hundredsImage.jpg')
print ('Hundreds place cropped, binarized, and saved')
def get_pan_info(img_path):
if img_path == None:
img_path = sys.argv[1]
print "Reading from file: " + img_path
img = Image(img_path).toRGB()
min_pan_area = img.width * img.height / 10
pan_rects = img.invert().findBlobs((25, 0, 0), min_pan_area)
if pan_rects and len(pan_rects) > 0:
img = img.crop(pan_rects[0])
if img.width < 350 or img.height < 200:
print "Error too small PAN image"
exit(1)
cropped_img = img.crop(0, img.height / 4.60, img.width / 1.4,
img.height / 4 * 2.3)
if cropped_img.width < 550:
cropped_img = cropped_img.resize(
550, 550 * cropped_img.height / cropped_img.width)
t_img = cropped_img.threshold(100)
img_text = pytesseract.image_to_string(t_img.getPIL(), lang='eng')
context = {}
learn_text(context, img_text)
t_img = cropped_img
img_text = pytesseract.image_to_string(t_img.getPIL(), lang='eng')
learn_text(context, img_text)
t_img = cropped_img.threshold(90)
img_text = pytesseract.image_to_string(t_img.getPIL(), lang='eng')
learn_text(context, img_text)
t_img = cropped_img.threshold(110)
img_text = pytesseract.image_to_string(t_img.getPIL(), lang='eng')
learn_text(context, img_text)
#print context['best_list']
pan_info = get_fields(context['best_list'])
#print json.dumps(pan_info, indent=2)
return pan_info
def blob(id):
cur = g.db.execute("SELECT id, image, x, y, width, height FROM annotations WHERE annotations.id=?", [id])
line = cur.fetchone()
img = Image("static\\img\\%d.jpg" % int(line[1]))
blob = img.crop(line[2], line[3], line[4], line[5])
io = StringIO()
blob.save(io)
data = io.getvalue()
resp = make_response(data)
resp.content_type = "image/jpeg"
return resp
def getImage(self):
if isinstance(self._scv_cam, Kinect):
if self._scv_cam._usedepth == 1:
img = self._scv_cam.getDepth()
elif self._scv_cam._usematrix == 1:
mat = self._scv_cam.getDepthMatrix().transpose()
img = Image(np.clip(mat - np.min(mat), 0, 255))
else:
img = self._scv_cam.getImage()
else:
img = self._scv_cam.getImage()
if self.crop:
img = img.crop(self.crop)
return img
def blob(id):
cur = g.db.execute(
'SELECT id, image, x, y, width, height FROM annotations WHERE annotations.id=?',
[id])
line = cur.fetchone()
img = Image("static\\img\\%d.jpg" % int(line[1]))
blob = img.crop(line[2], line[3], line[4], line[5])
io = StringIO()
blob.save(io)
data = io.getvalue()
resp = make_response(data)
resp.content_type = "image/jpeg"
return resp
def capture_camera_image():
log("capture_camera_image()")
image = None
if USE_MOTION_CAMERA:
image = Image(MOTION_CAMERA_SNAPSHOT)
else: # default to USE_WEB_CAMERA
if cam is None:
image = Camera().getImage()
else:
image = cam.getImage()
size = image.size()
# dynamic background mean color calculation
if (BACKGROUND_AUTO_CALIBRATE):
x = CROP_IMAGE_BORDERS[0]
y = CROP_IMAGE_BORDERS[1]
if (x > 0 and y > 0):
background = image.crop(0, 0, x, y)
mean_color = background.meanColor()
log("background as (%d, %d, %d)" % (mean_color[0], mean_color[1], mean_color[2]))
# save it
db = db_open()
db_backgrounds_write(db, db_background(date_created=datetime.now(), mean_color=mean_color))
db_close(db)
# if image border needs to be cropped
x = CROP_IMAGE_BORDERS[0]
y = CROP_IMAGE_BORDERS[1]
if (x > 0 and y > 0):
width = size[0] - 2 * x
height = size[1] - 2 * y
cropped = image.crop(x, y, width, height)
else:
cropped = image
return cropped
def line_blobs(id):
cur = g.db.execute("SELECT id, x, y, width, height FROM blobs WHERE image=?", [id])
blobs = cur.fetchall()
entries = []
img = Image("static\\img\\%d.jpg" % id)
for i, entry in enumerate(blobs):
blob = img.crop(entry[1], entry[2], entry[3], entry[4])
if blob and "SK_MODEL" in app.config:
if blob.height > blob.width:
blob = blob.resize(h=app.config["PATCH_SIZE"])
else:
blob = blob.resize(w=app.config["PATCH_SIZE"])
blob = blob.embiggen((app.config["PATCH_SIZE"], app.config["PATCH_SIZE"]))
np_img = blob.getGrayNumpy().transpose().reshape(-1)
pred = labels.inverse_transform(sk_model.predict(np_img))[0]
if app.config["DEBUG"]:
blob.save("tmp\\pic%d %s.jpg" % (i, pred))
entries.append(dict(x=entry[1], y=entry[2], width=entry[3], height=entry[4], pred=pred))
else:
entries.append(dict(x=entry[1], y=entry[2], width=entry[3], height=entry[4]))
return jsonify(blobs=entries)
def get_faces(image_path, api_mode=False, rescale_face_crop=0):
"""
Return a list of cropped faces given an image path
:param image_path: Path to image
:type image_path: str
:param api_mode: If api_mode is True get_faces returns a list of found HaarFeatures
:type api_mode: bool
:returns: list of images
"""
original_image = Image(image_path)
faces = original_image.findHaarFeatures(segment_face)
if api_mode:
return faces
else:
if rescale_face_crop:
return [original_image.crop(scale_bounding_box(face.boundingBox(), rescale_face_crop)) for face in faces]
else:
return [face.crop() for face in faces]
def generate_negative_examples(argv):
image_dirs = argv[4:]
images = []
for image_dir in image_dirs:
# grab all images
images.extend(glob(path.join(image_dir, '*.jpg')))
images.extend(glob(path.join(image_dir, '*.JPG')))
images.extend(glob(path.join(image_dir, '*.png')))
images.extend(glob(path.join(image_dir, '*.PNG')))
images = set(images)
if len(images) < N:
print 'Not enough images! (got %d, need %d)' % (len(images), N)
return
width, height, output_dir = int(argv[1]), int(argv[2]), argv[3]
if path.exists(output_dir) and (not path.isdir(output_dir)):
print '%s is not a directory' % output_dir
return
elif not path.exists(output_dir):
os.mkdir(output_dir)
for i in xrange(N):
print 'generating %3d/%d...' % ((i+1), N)
img = Image(images.pop())
img = img.grayscale()
if img.width > MAX_WIDTH:
img = img.resize(MAX_WIDTH, int(1.0*img.height*MAX_WIDTH/img.width))
x, y = random.randint(0, img.width-width), random.randint(0, img.height-height)
img = img.crop(x, y, width, height)
path_to_save = path.join(output_dir, '%d.png' % (i+1))
img.save(path_to_save)
def histo(self):
cam_mode = self.cam_mode
js = self.js
ms = self.ms
w = self.w
cent = 0
rgb1 = 0
c2 = self.c2
wsh = self.wsh
wsh2 = self.wsh2
s.write('s')
if cam_mode == 3:
img1 = c2.getImage()
if cam_mode==1:
with picamera.PiCamera() as camera:
camera.resolution = (544, 288)
camera.capture('imagesmall.jpg')
img1 = Image('imagesmall.jpg')
if cam_mode==2:
with picamera.PiCamera() as camera:
camera.resolution = (544, 288)
camera.capture('imagesmall.jpg')
img1 = Image('imagesmall.jpg')
self.img1 = img1
blobs = img1.findBlobs()
if blobs:
print "blob"
x = self.x
y = self.y
p = self.p
p = p + 1
img1.drawCircle((blobs[-1].x,blobs[-1].y),30,color=(255,255,255))
img1.drawCircle((blobs[0].centroid()),10,color=(255,100,100))
print blobs[-1].meanColor()
rgb1 = blobs[-1].meanColor()
cent = blobs[-1].centroid()
pth1 = "/var/www/images/image"
pth3 = ".png"
pth = pth1 + str(p) + pth3
print pth
img1.save(pth)
thumbnail = img1.crop(150,25,250,250)
thumbnail = thumbnail.scale(20,20)
thumb1 = "/var/www/images/thumbs/thumb"
thumb3 = ".png"
thumbpath = thumb1 + str(p) + thumb3
print thumbpath
thumbnail.save(thumbpath)
self.p = p
mySet.add((p,x,y,w,cent,rgb1))
self.mySet = mySet
wshx = str(self.x)
wshy = str(self.y)
centroidx = int(cent[0])
centroidy=int(cent[1])
rcolor=rgb1[0]
gcolor=rgb1[1]
bcolor=rgb1[2]
rcolor=int(rcolor)
gcolor=int(gcolor)
bcolor=int(bcolor)
wsh.write_message(wsh2, "rgb_" + str(rcolor)+"_"+str(gcolor)+"_"+str(bcolor))
wsh.write_message(wsh2, "x_" + str(centroidx)+"_"+str(centroidy))
img1.save(js.framebuffer)
wsh.write_message(wsh2, "d_" + wshx + "_" + wshy + "_" + str(p) )
else:
wshx = str(self.x)
wshy = str(self.y)
wsh.write_message(wsh2, wshx + " " + wshy + "dark")
print "dark"
from SimpleCV import Image
import time
img = Image('ladies.jpg')
# Crop starting at +(50, 5)+ for an area 280 pixels wide by 500 pixels tall
cropImg = img.crop(80, 5, 280, 500)
cropImg.show()
time.sleep(10)
#import the modules
import time
from SimpleCV import Image
from SimpleCV import Color
car_in_lot = Image("parking-car.png")
car_not_in_lot = Image("parking-no-car.png")
car = car_in_lot.crop(470, 200, 200, 200)
win = car.show()
time.sleep(1)
win.quit()
yellow_car = car.colorDistance(Color.YELLOW)
win = yellow_car.show()
time.sleep(1)
win.quit()
only_car = car - yellow_car
win = only_car.show()
time.sleep(1)
win.quit()
print('resultados primera comparacion:')
R = only_car.meanColor()[2]
print('R: ' + str(R))
G = only_car.meanColor()[1]
print('G: ' + str(G))
B = only_car.meanColor()[0]
print('B: ' + str(B))
def histo(self):
cam_mode = self.cam_mode
js = self.js
ms = self.ms
w = self.w
cent = 0
rgb1 = 0
c2 = self.c2
wsh = self.wsh
wsh2 = self.wsh2
s.write('s')
if cam_mode == 3:
img1 = c2.getImage()
if cam_mode == 1:
with picamera.PiCamera() as camera:
camera.resolution = (544, 288)
camera.capture('imagesmall.jpg')
img1 = Image('imagesmall.jpg')
if cam_mode == 2:
with picamera.PiCamera() as camera:
camera.resolution = (544, 288)
camera.capture('imagesmall.jpg')
img1 = Image('imagesmall.jpg')
self.img1 = img1
blobs = img1.findBlobs()
if blobs:
print "blob"
x = self.x
y = self.y
p = self.p
p = p + 1
img1.drawCircle((blobs[-1].x, blobs[-1].y),
30,
color=(255, 255, 255))
img1.drawCircle((blobs[0].centroid()), 10, color=(255, 100, 100))
print blobs[-1].meanColor()
rgb1 = blobs[-1].meanColor()
cent = blobs[-1].centroid()
pth1 = "/var/www/images/image"
pth3 = ".png"
pth = pth1 + str(p) + pth3
print pth
img1.save(pth)
thumbnail = img1.crop(150, 25, 250, 250)
thumbnail = thumbnail.scale(20, 20)
thumb1 = "/var/www/images/thumbs/thumb"
thumb3 = ".png"
thumbpath = thumb1 + str(p) + thumb3
print thumbpath
thumbnail.save(thumbpath)
self.p = p
mySet.add((p, x, y, w, cent, rgb1))
self.mySet = mySet
wshx = str(self.x)
wshy = str(self.y)
centroidx = int(cent[0])
centroidy = int(cent[1])
rcolor = rgb1[0]
gcolor = rgb1[1]
bcolor = rgb1[2]
rcolor = int(rcolor)
gcolor = int(gcolor)
bcolor = int(bcolor)
wsh.write_message(
wsh2,
"rgb_" + str(rcolor) + "_" + str(gcolor) + "_" + str(bcolor))
wsh.write_message(wsh2,
"x_" + str(centroidx) + "_" + str(centroidy))
img1.save(js.framebuffer)
wsh.write_message(wsh2, "d_" + wshx + "_" + wshy + "_" + str(p))
else:
wshx = str(self.x)
wshy = str(self.y)
wsh.write_message(wsh2, wshx + " " + wshy + "dark")
print "dark"
from SimpleCV import Image
img = Image('ahmedabad.jpg')
cropImg = img.crop(50, 5, 200, 200)
cropImg.show()
raw_input()
from __future__ import print_function
from SimpleCV import Image, Color
im = Image("img/ardrone.jpg")
# im = im.erode()
im = im.crop(2000, 1500, 2000, 2000, centered=True)
# im = im.erode()
# blobs = im.findBlobs()
# blobs.draw()
only_orange = im - im.colorDistance(Color.ORANGE)
only_black = im - im.colorDistance(Color.BLACK)
only_drone = only_orange + only_black
body.save("img/ardrone2.jpg")
# raw_input("Hit enter to quit: ")
# window.quit()
def cuteImg(pack_blob_zoom,pack_blob_size):
oriImg = Image("./img/original.jpg")
cropImg = oriImg.crop(pack_blob_zoom[0]*2,pack_blob_zoom[1]*2.05,pack_blob_size[0]*0.85,pack_blob_size[1]*0.58,centered=True)
cropImg.save("./img/dest5.jpg")
from SimpleCV import Image, Color, Display
import argparse
parser = argparse.ArgumentParser(description='Check for car in handicap spot.')
parser.add_argument('img')
args = parser.parse_args()
img = args.img
image = Image(img)
cropped_image = image.crop(470, 200, 200, 200)
color_distance = cropped_image.colorDistance(Color.YELLOW)
spot = cropped_image - color_distance
spot = spot.toRGB()
(r, g, b) = spot.meanColor()
if ((r > 15) and (g > 10)):
print "The car is in the lot. Call the attendant."
else:
print "The car is not in the lot."
from SimpleCV import Image, Color, Display
import argparse
parser = argparse.ArgumentParser(description='Check for car in handicap spot.')
parser.add_argument('img')
args = parser.parse_args()
img = args.img
image = Image(img)
cropped_image = image.crop(470,200,200,200)
color_distance = cropped_image.colorDistance(Color.YELLOW)
spot = cropped_image - color_distance
spot = spot.toRGB()
(r, g, b) =spot.meanColor()
if ((r>15) and (g>10)):
print "The car is in the lot. Call the attendant."
else:
print "The car is not in the lot."
#import the modules
import time
from SimpleCV import Image
from SimpleCV import Color
car_in_lot = Image("parking-car.png")
car_not_in_lot = Image("parking-no-car.png")
car = car_in_lot.crop(470,200,200,200)
win = car.show()
time.sleep(1)
win.quit()
yellow_car = car.colorDistance(Color.YELLOW)
win = yellow_car.show()
time.sleep(1)
win.quit()
only_car = car - yellow_car
win = only_car.show()
time.sleep(1)
win.quit()
print('resultados primera comparacion:')
R = only_car.meanColor()[2]
print('R: ' + str(R))
G = only_car.meanColor()[1]
print('G: ' + str(G))
B = only_car.meanColor()[0]
print('B: ' + str(B))
# time.sleep(5) ## Cropping the image into the four pieces ## The x coords are always the whole size x = 55 segment_height = img_size[1]/4 y_coords = [i * segment_height for i in range(4)] # print segment_height, y_coords avg_sizes = [] for y in y_coords: ## Crop the image (x, y, width, height) channel = droplets.crop(x, y, img_size[0], segment_height) # channel.show() ## Find droplets (blobs of color) blobs = channel.findBlobs() ## If blobs found if blobs: ## Keep only reasonably sized blobs blobs = [b for b in blobs if b.area() < 1000] ## Get the coordinates of each blob if it is reasonably sized points = [b.coordinates() for b in blobs] ## Draw the center point channel.drawPoints(points, color=(255, 0, 0), sz=3, width=-1)
def fancify():
if request.method == 'POST':
print request.data
cur_request = json.loads(request.data)
else:
#cur_request = """{"url": "", "debug":true}"""
#cur_request = """{"url": "", "debug":true}"""
#cur_request = """{"url": "", "debug":true}"""
cur_request = """{"url": "http://localhost/images/scrubs.jpg", "debug":true}"""
#cur_request = """{"url": "http://www.newrichstrategies.com/wp-content/uploads/2012/03/How-to-Find-Good-People-in-Your-Life.jpg", "debug":false}"""
#cur_request = """{"url": "http://greenobles.com/data_images/frank-lampard/frank-lampard-02.jpg", "debug":true}"""
#cur_request = """{"url": "http://www.billslater.com/barack__obama.jpg"}"""
#cur_request = """{"url": "http://celebrityroast.com/wp-content/uploads/2011/01/arnold-schwarzenegger-body-building.jpg", "debug":false}"""
#cur_request = """{"url": "http://face2face.si.edu/.a/6a00e550199efb8833010536a5483e970c-800wi", "debug":true}"""
#cur_request = """{"url": "http://collider.com/uploads/imageGallery/Scrubs/scrubs_cast_image__medium_.jpg", "debug":false}"""
#cur_request = """{"url": "http://localhost/images/Kevin_Bacon_at_the_2010_SAG_Awards.jpg", "debug":false}"""
#cur_request = """{"url": "http://cdn02.cdn.justjared.com/wp-content/uploads/headlines/2012/02/anna-faris-oscars-red-carpet-2012.jpg", "debug":true}"""
#cur_request = """{"url": "http://www.viewzone.com/attractive.female.jpg", "debug":true}"""
cur_request = json.loads(cur_request)
print cur_request["url"]
img = Image(str(cur_request["url"]))
img = img.scale(2.0)
debug = True
#if "debug" in cur_request:
# debug = cur_request["debug"]
chosen_faces = []
faces = img.findHaarFeatures(face_cascade)
if faces is not None:
for face in faces:
face_features = []
invalid_face = False
face_rect = Rect(face.x - (face.width() / 2), face.y - (face.height() / 2), face.width(), face.height())
for chosen_face in chosen_faces:
if face_rect.colliderect(chosen_face):
invalid_face = True
break
if invalid_face:
break
nose = None
mouth = None
left_eye = None
right_eye = None
cur_face = img.crop(face.x, face.y, face.width(), face.height(), centered=True)
#cur_face = face.crop()
noses = cur_face.findHaarFeatures(nose_cascade)
mouths = cur_face.findHaarFeatures(mouth_cascade)
eyes = cur_face.findHaarFeatures(eye_cascade)
face_left_edge = face.x - (face.width() / 2)
face_top_edge = face.y - (face.height() / 2)
if noses is not None:
nose = noses[0]
nose_dist = (abs(nose.x - (face.width() / 2)) +
abs(nose.y - (face.height() * 5 / 9)) +
abs(nose.width() - (face.width() / 4)))
for cur_nose in noses:
cur_dist = (abs(cur_nose.x - (face.width() / 2)) +
abs(cur_nose.y - (face.height() * 5 / 9)) +
abs(cur_nose.width() - (face.width() / 4)))
if cur_dist < nose_dist:
nose = cur_nose
nost_dist = cur_dist
if nose and (nose.y < (face.height() / 3)):
nose = None
if nose and mouths is not None:
mouth = mouths[0]
mouth_dist = abs(mouth.x - nose.x) + (abs(mouth.y - (face.height() * 4 / 5)) * 2)
for cur_mouth in mouths:
cur_dist = abs(cur_mouth.x - nose.x) + (abs(cur_mouth.y - (face.height() * 4/ 5)) * 2)
if (cur_dist < mouth_dist) and (cur_mouth.y > nose.y):
mouth = cur_mouth
mouth_dist = cur_dist
if nose and eyes:
right_eye = eyes[0]
right_eye_dist = (abs(right_eye.x - (3 * face.width() / 4)) * 2 +
abs(right_eye.y - (nose.y - (nose.height() / 2)) / 2) +
abs(right_eye.width() - (face.width() / 3)))
for cur_eye in eyes:
cur_right_dist = (abs(cur_eye.x - (3 * face.width() / 4)) +
abs(cur_eye.y - (nose.y - (nose.height() / 2)) / 2) +
abs(cur_eye.width() - (face.width() / 3)))
if (cur_right_dist <= right_eye_dist): # and (cur_eye.y < nose.y):
right_eye = cur_eye
right_eye_dist = cur_right_dist
if nose and right_eye and (((right_eye.y - (right_eye.height() / 2)) > nose.y) or (right_eye.x < nose.x)):
print "Culling right_eye"
right_eye = None
if nose and mouth:
chosen_faces.append(face_rect)
x_face = face.x - (face.width() / 2)
y_face = face.y - (face.height() / 2)
x_nose = nose.x - (nose.width() / 2)
y_nose = nose.y - (nose.height() / 2)
# Setup TopHat Image
scale_factor = face.width() / 175.0
cur_hat = hat.copy()
cur_hat = cur_hat.scale(scale_factor)
cur_hat_mask = hat_mask.copy()
cur_hat_mask = cur_hat_mask.scale(scale_factor)
cur_hat_mask = cur_hat_mask.createAlphaMask(hue_lb=0, hue_ub=100)
# Calculate the hat position
if (face.y - face.height() / 2) > cur_hat.height:
x_hat = face.x - (cur_hat.width / 2)
y_hat = face.y - (face.height() * 7 / 10) - (cur_hat.height / 2)
img = img.blit(cur_hat, pos=(x_hat, y_hat), alphaMask=cur_hat_mask)
if mouth:
x_mouth = mouth.x - (mouth.width() / 2)
y_mouth = mouth.y - (mouth.height() / 2)
# Setup Mustache Image
cur_stache = stache.copy()
scale_factor = ((nose.width() / 300.0) + (face.width() / 600.0)) / 2.0
cur_stache = cur_stache.scale(scale_factor)
stache_mask = cur_stache.createAlphaMask(hue_lb=0, hue_ub=10).invert()
# Calculate the mustache position
bottom_of_nose = y_nose + (nose.height() * 4 / 5)
top_of_mouth = y_mouth
# if top_of_mouth > bottom_of_nose:
# top_of_mouth = bottom_of_nose
y_must = y_face + ((bottom_of_nose + top_of_mouth) / 2) - (cur_stache.height / 2)
middle_of_nose = nose.x
middle_of_mouth = mouth.x
x_must = x_face + ((middle_of_nose + middle_of_mouth) / 2) - (cur_stache.width / 2)
if right_eye:
x_right_eye = right_eye.x - (right_eye.width() / 2)
y_right_eye = right_eye.y - (right_eye.height() / 2)
# Setup Monocle Image
cur_mono = monocle.copy()
scale_factor = ((right_eye.width() / 65.0) + (face.width() / 200.0)) / 2.0
cur_mono = cur_mono.scale(scale_factor)
mono_mask = cur_mono.createAlphaMask(hue_lb=0, hue_ub=100).invert()
# Calculate Monocle Position
x_mono = x_face + x_right_eye
y_mono = y_face + y_right_eye
img = img.blit(cur_mono, pos=(x_mono, y_mono), alphaMask=mono_mask)
img = img.blit(cur_stache, pos=(x_must, y_must), alphaMask=stache_mask)
if debug:
noselayer = DrawingLayer((img.width, img.height))
nosebox_dimensions = (nose.width(), nose.height())
center_point = (face.x - (face.width() / 2) + nose.x,
face.y - (face.height() / 2) + nose.y)
nosebox = noselayer.centeredRectangle(center_point, nosebox_dimensions, width=3)
img.addDrawingLayer(noselayer)
img = img.applyLayers()
else:
print "Face culled:"
if not nose:
print " No Nose"
if not mouth:
print " No mouth"
if not right_eye:
print " No right eye"
print
if debug:
face_left_edge = face.x - (face.width() / 2)
face_top_edge = face.y - (face.height() / 2)
facelayer = DrawingLayer((img.width, img.height))
facebox_dimensions = (face.width(), face.height())
center_point = (face.x, face.y)
facebox = facelayer.centeredRectangle(center_point, facebox_dimensions, Color.BLUE)
img.addDrawingLayer(facelayer)
if noses:
for nose in noses:
noselayer = DrawingLayer((img.width, img.height))
nosebox_dimensions = (nose.width(), nose.height())
center_point = (face.x - (face.width() / 2) + nose.x,
face.y - (face.height() / 2) + nose.y)
nosebox = noselayer.centeredRectangle(center_point, nosebox_dimensions)
img.addDrawingLayer(noselayer)
if mouths:
for mouth in mouths:
mouthlayer = DrawingLayer((img.width, img.height))
mouthbox_dimensions = (mouth.width(), mouth.height())
center_point = (face.x - (face.width() / 2) + mouth.x,
face.y - (face.height() / 2) + mouth.y)
mouthbox = mouthlayer.centeredRectangle(center_point, mouthbox_dimensions, Color.GREEN)
img.addDrawingLayer(mouthlayer)
if eyes:
for right_eye in eyes:
right_eyelayer = DrawingLayer((img.width, img.height))
right_eyebox_dimensions = (right_eye.width(), right_eye.height())
right_eye_center_point = (face_left_edge + right_eye.x, face_top_edge + right_eye.y)
right_eyebox = right_eyelayer.centeredRectangle(right_eye_center_point, right_eyebox_dimensions)
img.addDrawingLayer(right_eyelayer)
img = img.applyLayers()
img = img.scale(0.5)
w_ratio = img.width / 800.0
h_ratio = img.height / 600.0
if h_ratio > 1.0 or w_ratio > 1.0:
if h_ratio > w_ratio:
img = img.resize(h=600)
else:
img = img.resize(w=800)
output = StringIO.StringIO()
img.getPIL().save(output, format="JPEG") #, quality=85, optimize=True)
img_contents = output.getvalue()
mimetype = "image/jpeg"
return app.response_class(img_contents, mimetype=mimetype, direct_passthrough=False)
from __future__ import print_function
from SimpleCV import Image, Color, VirtualCamera, Display
import SimpleCV as scv
# # video = VirtualCamera('', 'video')
# display = Display()
# while display.isNotDone():
# img = video.getImage()
# try:
# dist = img - img.colorDistance(Color.RED)
# dist.show()
# except KeyboardInterrupt:
# display.done = True
# if display.mouseRight:
# display.done = True
# display.quit()
img = Image('img/curb.JPG')
img = img.crop(0, 2*img.height/3, img.width, 5*img.height/6)
print(img.meanColor())
# img = img.binarize()
# img.findBlobs().draw()
# ls = img.findLines()
# for l in ls:
# l.draw()
# img.findBlobs().draw()
# img = img.colorDistance(Color.RED)
img.save('img/curb_processed.jpg')
from SimpleCV import Image, Color, Display
import time
car_in_lot = Image("parking-car.png")
car = car_in_lot.crop(470, 200, 200, 200)
yellow_car = car.colorDistance(Color.YELLOW)
only_car = car - yellow_car
only_car = only_car.toRGB()
displayObject = Display()
print only_car.meanColor()
# Show the results
only_car.save(displayObject)
while displayObject.isNotDone():
time.sleep(0.5)
from SimpleCV import Image,Color
car_in_lot = Image("parking-car.png")
car_not_in_lot = Image("parking-no-car.png")
car = car_in_lot.crop(470,200,200,200)
yellow_car = car.colorDistance(Color.YELLOW)
yellow_car.show()
img = Image("SquareCube1.jpg") # Obtain image with cube within the red square
test = {} # Dictionary containing difference images of a particular sticker and standard colors
pixValues = [] # List containing the pixel values from (20,50) to (20,80) of each image present in test
averagesDict = {} # Dictionary containing the average of pixValues corresponding to each color
averagesList = [] # List containing all the average of pixValues corresponding to each color
stickerColors = [[],[],[]] # Contains first letter of color of all the stickers in a picture
# Coordinates corresponding to the 'origin' of the cube
squareCornerX = 170
squareCornerY = 90
for i in range(0,3):
for j in range(0,3):
croppedImg = img.crop(squareCornerX + j*100, squareCornerY + i*100, 100, 100) # Obtain just the sticker
# Find the difference image of the sticker and all the standard colors and store it in test
for color in rubikColor:
test[color] = croppedImg.colorDistance(color = rubikColor[color])
# Obtain values of a line in the middle of the difference image
# From observation, we see best results with pixels from (20,50) to (80,50)
for j1 in range(20,81):
pixValues.append(test[color].getPixel(j1,50)[0])
average = np.average(pixValues) # Find average of pixel values
averagesList.append(average) # And add the value to averagesList
averagesDict[color] = average # And also add the value to averagesDict
pixValues = [] # Clear pixValues
from SimpleCV import Image
img = Image('ex16.png')
# Crop starting at +(50, 5)+ for an area 400 pixels wide by 400 pixels tall
cropImg = img.crop(50, 5, 400, 400)
cropImg.show()
def histo(self): ## this def is the "light meter" part---
cam_mode = self.cam_mode ## the pic gets cataloged if true ---
js = self.js
w = self.w
cent = 0
rgb1 = 0
c2 = self.c2
wsh = self.wsh
wsh2 = self.wsh2
i=0
brightpixels=0
darkpixels=0
blobs = 0
if cam_mode == 3: ## sort out the confusing cam modes
img1 = c2.getImage()
time.sleep(.25)
if cam_mode==1:
with picamera.PiCamera() as camera:
camera.resolution = (544, 288)
camera.capture('imagesmall.jpg')
img1 = Image('imagesmall.jpg')
if cam_mode==2:
with picamera.PiCamera() as camera:
camera.resolution = (544, 288)
camera.capture('imagesmall.jpg')
img1 = Image('imagesmall.jpg')
blobs = img1.findBlobs()
time.sleep(0.5)
if blobs:
##find the blob centroid and cut it out 20x20
crop1 = blobs[-1].x
crop2 = blobs[-1].y
crop3 = crop1 - 10
crop4 = crop2 - 10
thumbnail = img1.crop(crop3,crop4,20,20)
img2 = thumbnail
hist = img2.histogram(20)
## split the thumb into 20 levels of darkness
brightpixels = hist[10]
## 10 is where the darkest of the light pixels accumulate
print brightpixels
##while i < 20:
##old code for if you want to split the histogram in two
##if (i < 10):
##darkpixels = darkpixels + hist[i]
##self.darkpixels = darkpixels
##print hist[i]
##else:
##brightpixels = brightpixels + hist[i]
##self.brightpixels = brightpixels
##print hist[i]
##i = i + 1
if (brightpixels<400): ## heres where it decides to catalog the pic or not...
wsh.write_message(wsh2, "histo_" + str(darkpixels) + "_" + str(brightpixels))
print "blob"
x = self.x
y = self.y
p = self.p
p = p + 1
thumb1 = "/var/www/html/images/thumbs/thumb"
thumb3 = ".png"
thumbpath = thumb1 + str(p) + thumb3
print thumbpath
thumbnail.save(thumbpath)
img1.drawText("blob = True", 10, 35, color=(255,255,255),fontsize=30)
img1.drawText("search_mode", 10, 5, color=(0,0,255),fontsize=40)
img1.drawText("blob centroid", blobs[-1].x,blobs[-1].y, color=(255,255,255),fontsize=20)
img1.drawCircle((blobs[-1].x,blobs[-1].y),30,color=(255,255,0))
img1.drawCircle((blobs[0].centroid()),10,color=(255,255,255))
print blobs[-1].meanColor()
rgb1 = blobs[-1].meanColor()
cent = blobs[-1].centroid()
pth1 = "/var/www/html/images/image"
pth3 = ".png"
pth = pth1 + str(p) + pth3
print pth
img1.save(pth)
time.sleep(0.5)
self.p = p
mySet.add((p,x,y,w,cent,rgb1))
self.mySet = mySet
wshx = str(self.x)
wshy = str(self.y)
centroidx = int(cent[0])
centroidy=int(cent[1])
rcolor=rgb1[0]
gcolor=rgb1[1]
bcolor=rgb1[2]
rcolor=int(rcolor)
gcolor=int(gcolor)
bcolor=int(bcolor)
wsh.write_message(wsh2, "rgb_" + str(rcolor)+"_"+str(gcolor)+"_"+str(bcolor))
wsh.write_message(wsh2, "x_" + str(centroidx)+"_"+str(centroidy))
img1.save(js.framebuffer)
time.sleep(0.5)
wsh.write_message(wsh2, "d_" + wshx + "_" + wshy + "_" + str(p) )
else:
print "senosor dark"
print darkpixels,'_', brightpixels
blobs = 0
wsh.write_message(wsh2, "histo_" + str(darkpixels) + "_" + str(brightpixels))
else:
wshx = str(self.x)
wshy = str(self.y)
wsh.write_message(wsh2, wshx + " " + wshy + "dark")
img1.save(js.framebuffer)
print "no blob"
