def impose(id, overlay, px, py):
img = saved_vars.get_var(id)
other_image = saved_vars.get_var(overlay)
# Mask ranges from 0 to width / height of overlay
mask = np.zeros(img.shape, dtype=np.bool)
mask[:other_image.shape[0], :other_image.shape[1]] = True
locs = np.where(mask != 0) # Get the non-zero mask locations
# Following conditional logic is equivalent to copyTo from other languages
# TODO implement overflow logic to cutoff instead of wrap
# Background is colored but overlay is grayscale
if len(img.shape) == 3 and len(other_image.shape) != 3:
img[locs[0] + px, locs[1] + py] = other_image[locs[0], locs[1], None]
# Both overlay and background are grayscale
elif (len(img.shape) == 3 and len(other_image.shape) == 3) or \
(len(img.shape) == 1 and len(other_image.shape) == 1):
img[locs[0] + px, locs[1] + py] = other_image[locs[0], locs[1]]
# Otherwise, we can't do this
else:
raise hiphop_error("InvalidFunctionError",
"Incompatible input and output dimensions.")
saved_vars.add_var(id, img, saved_vars.get_path(id))
return
def outline(id, value):
# print("Outline function called. Paramters: {}, {}".format(id, value))
kernel = np.ones((value,value), np.uint8)
img = saved_vars.get_var(id)
morph_gradient = cv2.morphologyEx(img, cv2.MORPH_GRADIENT, kernel)
saved_vars.add_var(id, morph_gradient)
return
def dilate(id, value):
# print("Dilating function called. Parameters: {}, {}".format(id, value))
kernel = np.ones((value,value),np.uint8)
img = saved_vars.get_var(id)
eroded = cv2.dilate(img,kernel,iterations = 1)
saved_vars.add_var(id, eroded)
return
def savefile(id, filename):
# print("Save file function. Parameters: {}, {}".format(id, filename))
if (filename.startswith('../')):
# should throw error here
raise file_error("SaveError", "Cannot save file, filename should not start with ../")
new_filename = filename
# id filename just start with /
if (filename.startswith('/')):
new_filename = "." + filename
# if filename does not start with ./
if not (new_filename.startswith('./')):
new_filename = "./" + filename
head_tail = os.path.split(new_filename)
if not (os.path.exists(head_tail[0])):
os.makedirs(head_tail[0])
cv2.imwrite(new_filename, saved_vars.get_var(id))
print(colored("Image at id {} successfully saved to {}".format(id, filename), "green"))
def grayscale(id):
# print("Grayscale function called. Parameters: {}".format(id))
img = saved_vars.get_var(id)
gray_image = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
saved_vars.add_var(id, gray_image)
return
def scale(id, x, y):
img = saved_vars.get_var(id)
width = int(img.shape[1] * x)
height = int(img.shape[0] * y)
dim = (width,height)
scaled = cv2.resize(img, dim)
saved_vars.add_var(id, scaled)
return
def crop(id, widthlow, widthhigh, heightlow, heighthigh):
img = saved_vars.get_var(id)
height, width, channels = img.shape
heighthalf = height/2
widthhalf = width/2
hl = round(heighthalf + heightlow * heighthalf)
hh = round(heighthalf + heighthigh * heighthalf)
wl = round(widthhalf + widthlow * widthhalf)
wh = round(widthhalf + widthhigh * widthhalf)
crop_img = img[hl:hh, wl:wh]
saved_vars.add_var(id, crop_img)
return
def filtercolor(id, lowR, lowG, lowB, highR, highG, highB):
img = saved_vars.get_var(id)
hsv = cv2.cvtColor(img, cv2.COLOR_BGR2HSV)
# define range of color in HSV
lower_range = np.array([lowB, lowG, lowR])
upper_range = np.array([highB, highG, highR])
# Threshold the HSV image to get only specified colors
mask = cv2.inRange(hsv, lower_range, upper_range)
# Bitwise-AND mask and original image
res = cv2.bitwise_and(img ,img, mask=mask)
saved_vars.add_var(id, res)
return
def wave(id, direction, amplitude):
img = saved_vars.get_var(id)
img_output = np.zeros(img.shape, dtype=img.dtype)
rows, cols, mask = img.shape
if direction == "v":
# print("Dir v")
for i in range(rows):
for j in range(cols):
offset_x = int(int(amplitude) * np.sin(2 * 3.14 * i / 180))
offset_y = 0
if j + offset_x < rows:
img_output[i, j] = img[i, (j + offset_x) % cols]
else:
img_output[i, j] = 0
elif direction == "h":
# print("Dir h")
for i in range(rows):
for j in range(cols):
offset_x = 0
offset_y = int(int(amplitude) * np.sin(2 * 3.14 * j / 150))
if i + offset_y < rows:
img_output[i, j] = img[(i + offset_y) % rows, j]
else:
img_output[i, j] = 0
elif direction == "m":
for i in range(rows):
for j in range(cols):
offset_x = int(int(amplitude) * np.sin(2 * 3.14 * i / 150))
offset_y = int(int(amplitude) * np.cos(2 * 3.14 * j / 150))
if i + offset_y < rows and j + offset_x < cols:
img_output[i, j] = img[(i + offset_y) % rows,
(j + offset_x) % cols]
else:
img_output[i, j] = 0
else:
raise hiphop_error("InvalidFunctionError", "Invalid parameters.")
saved_vars.add_var(id, img_output, saved_vars.get_path(id))
return
def blur(id, value):
img = saved_vars.get_var(id)
blurred = cv2.blur(img, (value, value))
saved_vars.add_var(id, blurred)