def laplace_8(image):
g = np.array([[0, -1, 0],
[-1, 4, -1],
[0, -1, 0]])
h = conv(image, g)
h = np_value_range(h)
h = normalize(h)
show(h, "h")
save("laplace8-fig")
return h, "laplace8.jpg"
def roberts(image):
gx = np.array([[1, 0],
[0, -1]])
gy = np.array([[0, 1],
[-1, 0]])
hx = conv(image, gx)
hy = conv(image, gy)
h = np.abs(hx) + np.abs(hy)
h = np_value_range(h)
h = normalize(h)
show(hx, "Roberts-hx")
show(hy, "Roberts-hy")
show(h, "Roberts-h")
save("roberts-fig")
return h, "roberts.jpg"
def scharr(image):
gx = np.array([[-3, -10, -3],
[0, 0, 0],
[3, 10, 3]])
gy = np.array([[-3, 0, 3],
[-10, 0, 10],
[-3, 0, 3]])
hx = conv(image, gx)
hy = conv(image, gy)
h = np.sqrt(np.power(hx, 2.0) + np.power(hy, 2.0))
h = np_value_range(h)
h = normalize(h)
show(hx, "Scharr-hx")
show(hy, "Scharr-hy")
show(h, "Scharr-h")
save("scharr-fig")
return h, "scharr.jpg"
def sobel(image):
gx = np.array([[-1, -2, -1],
[0, 0, 0],
[1, 2, 1]])
gy = np.array([[-1, 0, 1],
[-2, 0, 2],
[-1, 0, 1]])
hx = conv(image, gx)
hy = conv(image, gy)
h = np.sqrt(np.power(hx, 2.0) + np.power(hy, 2.0))
h = np_value_range(h)
h = normalize(h)
show(hx, "Sobel-hx")
show(hy, "Sobel-hy")
show(h, "Sobel-h")
save("Sobel-fig")
return h, "sobel.jpg"
def prewitt(image):
gx = np.array([[-1, -1, -1],
[0, 0, 0],
[1, 1, 1]])
gy = np.array([[-1, 0, 1],
[-1, 0, 1],
[-1, 0, 1]])
hx = conv(image, gx)
hy = conv(image, gy)
h = np.sqrt(np.power(hx, 2.0) + np.power(hy, 2.0))
h = np_value_range(h)
h = normalize(h)
show(hx, "Prewitt-hx")
show(hy, "Prewitt-hy")
show(h, "Prewitt-h")
save("prewitt-fig")
return h, "prewitt.jpg"
def canny(image, high_thres=90, low_thres=30):
show(image, "origin")
height = len(image)
width = len(image[0])
# gauss filter
gauss = np.zeros((5, 5))
for i in range(5):
for j in range(5):
x = i - 2
y = j - 2
sigma = 1
gauss[i][j] = np.exp(-(x * x + y * y) /
(2 * sigma * sigma)) / \
(2 * np.arccos(-1) * sigma * sigma)
image = conv(image, gauss)
show(image, "after gauss")
# Sobel operator
gx = np.array([[-1, -2, -1],
[0, 0, 0],
[1, 2, 1]])
gy = np.array([[-1, 0, 1],
[-2, 0, 2],
[-1, 0, 1]])
hx = conv(image, gx)
hy = conv(image, gy)
h = np.sqrt(np.power(hx, 2.0) + np.power(hy, 2.0))
h = np_value_range(h)
h = normalize(h)
show(h, "calc gradient")
# maximum filter
res = np.zeros((height, width))
for i in range(1, height - 1):
for j in range(1, width - 1):
hx_ij = hx[i][j]
hy_ij = hy[i][j]
# up, down can be exchanged
up = -1
down = -1
if hx_ij == 0: # [90 270]
up = h[i - 1][j]
down = h[i + 1][j]
elif hy_ij == 0: # [0 180]
up = h[i][j + 1]
down = h[i][j - 1]
else:
if hy_ij <= 0:
hx_ij = -hx_ij
hy_ij = -hy_ij
if hx_ij > hy_ij: # [0, 45]
c = hy_ij / hx_ij
up = h[i][j + 1] * c + (1 - c) * h[i - 1][j + 1]
down = h[i][j - 1] * c + (1 - c) * h[i + 1][j - 1]
elif 0 < hx_ij < hy_ij: # [45, 90]
c = 1 - 1 / (hy_ij / hx_ij)
up = h[i - 1][j + 1] * c + (1 - c) * h[i - 1][j]
down = h[i + 1][j - 1] * c + (1 - c) * h[i + 1][j]
elif 0 < -hx_ij < hy_ij: # [90, 135]:
c = 1 / (-hy_ij / hx_ij)
up = h[i - 1][j] * c + (1 - c) * h[i - 1][j - 1]
down = h[i + 1][j] * c + (1 - c) * h[i + 1][j + 1]
else: # [135, 180]
c = 1 - (-hy_ij / hx_ij)
up = h[i - 1][j - 1] * c + (1 - c) * h[i][j - 1]
down = h[i + 1][j + 1] * c + (1 - c) * h[i][j + 1]
if h[i][j] >= up and h[i][j] >= down:
res[i][j] = h[i][j]
else:
res[i][j] = 0
show(res, "filter")
# 2-threshold filter
final = np.zeros((height, width))
for i in range(height):
for j in range(width):
if res[i][j] >= high_thres:
final[i][j] = 255
elif res[i][j] <= low_thres:
final[i][j] = 0
else:
final[i][j] = 1
change = True
previous = final.copy()
show(final, "before linking")
while change:
change = False
for i in range(height):
for j in range(width):
if previous[i][j] == 1:
states = [((previous[_i][_j] == 255) for _i in range(i - 1, i + 2)) for _j in range(j - 1, j + 2)]
if True in states:
final[i][j] = 255
change = True
else:
final[i][j] = previous[i][j]
show(final, "after linking")
show(final, "after linking")
save("canny-fig")
return final, "canny.jpg"