def basic_stimulus_null(inputfilen, outputfilen):
with open(inputfilen, 'rb') as input:
st = pickle.load(input).movie
sw2 = rh.RH(inputframe=st[..., 0])
rharr = np.empty_like(st)
for i in range(0, 120):
rharr[..., i] = sw2.run(st[..., 120 - i])
print('frame', i)
rharr = rharr * 0.05
plt.xlabel('frame')
plt.ylabel('response')
plt.plot(st[40, 40, 1:120], color='black', label='Image')
plt.plot(rharr[40, 40, 1:120], color='blue', label='Reichardt')
plt.legend(loc='best')
plt.savefig(outputfilen)
plt.clf()
def basic_stimulus_null(inputfilen, outputfilen):
with open(inputfilen, 'rb') as input:
st = pickle.load(input).movie
sw2 = rh.RH(
inputframe=st[..., 0],
A1_sigma=5,
A2_sigma=5,
)
rharr = np.empty_like(st)
A2B1 = np.empty_like(st)
A1B2 = np.empty_like(st)
for i in range(0, 120):
rharr[..., i] = sw2.run(st[..., 120 - i])
A2B1[..., i] = sw2.A2B1
A1B2[..., i] = sw2.A1B2
print('frame', i)
rharr = rharr * 1
A2B1 = A2B1 * 1
A1B2 = A1B2 * 1
fig, ax1 = plt.subplots()
fig.subplots_adjust(right=0.88)
ax2 = ax1.twinx()
ax1.set_ylim([-2, 2])
p1, = ax1.plot(st[40, 40, 1:120], color='black', label='Image')
p2, = ax2.plot(rharr[40, 40, 1:120], color='blue', label='Reichardt')
p3, = ax2.plot(A2B1[40, 40, 1:120], '--', color='green', label='A2B1')
p4, = ax2.plot(A1B2[40, 40, 1:120], '--', color='red', label='A1B2')
ax1.set_xlabel('frame')
ax1.set_ylabel('image intensity', color='black')
ax2.set_ylabel('response', color='blue')
lines = [p1, p2, p3, p4]
labs = [l.get_label() for l in lines]
ax1.legend(lines, labs, loc='best')
plt.savefig(outputfilen)
plt.clf()
img = cv2.add(Image, mask)
#print image
cv2.imshow(name, img)
return []
def main():
pass
if __name__ == '__main__':
foldername = './gabor/Yosemite/'
img = cv2.imread(foldername + 'frame07.png')
img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
sw2_h = rh.RH(inputframe=img, A1_Lambda=5.0, A2_Lambda=5.0)
sw2_v = rh.RH(inputframe=img,
A1_theta=0,
A2_theta=0,
A1_Lambda=5.0,
A2_Lambda=5.0)
# rharr = np.empty_like(st)
norm = mpl.colors.Normalize(vmin=-200, vmax=200)
for frame in range(7, 15):
img = cv2.imread(foldername + 'frame{0:02d}.png'.format(frame))
img00 = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
cv2.imshow('ss', img)
k = cv2.waitKey(0) & 0xff
if k == 27:
plt.legend(loc='best')
plt.savefig(outputfilen)
plt.clf()
def main():
pass
if __name__ == '__main__':
foldername = './gabor/Yosemite/'
img = cv2.imread(foldername + 'frame07.png')
img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
fheight, fwidth = img.shape
sw2 = rh.RH(inputframe=img, A1_Lambda=10.0, A2_Lambda=10.0)
sw3 = rh.RH(inputframe=img,
A1_theta=0,
A2_theta=0,
A1_Lambda=10.0,
A2_Lambda=10.0)
# rharr = np.empty_like(st)
norm = mpl.colors.Normalize(vmin=-40, vmax=40)
# sw2.print_gabor('gh0', 'gh1')
# sw3.print_gabor('gv0', 'gv1')
for frame in range(7, 15):
img = cv2.imread(foldername + 'frame{0:02d}.png'.format(frame))
imgb = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
flowx = sw2.run(imgb) * 0.00005
def basic_stimulus(inputfilen, outputfilen, preferdir=True, plot=True):
with open(inputfilen, 'rb') as input:
st = pickle.load(input).movie
sw2 = rh.RH(
inputframe=st[..., 0],
# A1_phase = 'cos',
# A2_phase = 'sin',
A1_sigma=5,
A2_sigma=5,
# A1_Lambda = 4.0,
# A2_Lambda = 4.0,
fl=0.25,
fh=0.25,
# winsize = 2
)
# sw2.print_gabor()
rharr = np.empty_like(st)
A2B1 = np.empty_like(st)
A1B2 = np.empty_like(st)
A1 = np.empty_like(st)
A2 = np.empty_like(st)
B1 = np.empty_like(st)
B2 = np.empty_like(st)
# sw2.print_gabor('gab3')
frame_range = 120
if (preferdir):
for i in range(0, frame_range):
rharr[..., i] = sw2.run(st[..., i])
A2B1[..., i] = sw2.A2B1
A1B2[..., i] = sw2.A1B2
A1[..., i] = sw2.convA1
A2[..., i] = sw2.convA2
B1[..., i] = sw2.lowpassA1
B2[..., i] = sw2.lowpassA2
print('frame', i)
else:
for i in range(0, frame_range):
rharr[..., i] = sw2.run(st[..., frame_range - i])
A2B1[..., i] = sw2.A2B1
A1B2[..., i] = sw2.A1B2
A1[..., i] = sw2.convA1
A2[..., i] = sw2.convA2
B1[..., i] = sw2.lowpassA1
B2[..., i] = sw2.lowpassA2
print('frame', i)
rharr = rharr * 1
A2B1 = A2B1 * 1
A1B2 = A1B2 * 1
fig, ax1 = plt.subplots()
fig.subplots_adjust(right=0.88)
ax2 = ax1.twinx()
ax1.set_ylim([-2, 2])
ax2.set_ylim([-45, 45])
sampleloc = 45
p1, = ax1.plot(st[sampleloc, sampleloc, 1:frame_range],
color='black',
label='input image')
# p2, = ax2.plot(rharr[sampleloc,sampleloc,1:frame_range], color='blue', label='Reichardt')
# p3, = ax2.plot(A2B1[sampleloc,sampleloc,1:frame_range], '--', color='green', label='A2B1')
# p4, = ax2.plot(A1B2[sampleloc,sampleloc,1:frame_range], '--', color='red', label='A1B2')
p5, = ax2.plot(A1[sampleloc, sampleloc, 1:frame_range],
color='green',
label='A1')
p6, = ax2.plot(A2[sampleloc, sampleloc, 1:frame_range],
color='red',
label='A2')
p7, = ax2.plot(B1[sampleloc, sampleloc, 1:frame_range],
'--',
color='green',
label='B1')
p8, = ax2.plot(B2[sampleloc, sampleloc, 1:frame_range],
'--',
color='red',
label='B2')
ax1.set_xlabel('frame number')
ax1.set_ylabel('image intensity', color='black')
ax2.set_ylabel('response', color='blue')
lines = [p1, p5, p6, p7, p8]
labs = [l.get_label() for l in lines]
ax1.legend(lines, labs, loc='best')
if (plot):
plt.savefig(outputfilen + '.png')
else:
plt.show()
plt.clf()
fig2, ax3 = plt.subplots()
fig2.subplots_adjust(right=0.88)
sampleloc = 45
# p1, = ax3.plot(st[sampleloc,sampleloc,1:frame_range], color='black', label='Input image')
p3, = ax3.plot(A2B1[sampleloc, sampleloc, 1:frame_range],
'--',
color='green',
label='A2B1')
p4, = ax3.plot(A1B2[sampleloc, sampleloc, 1:frame_range],
'--',
color='red',
label='A1B2')
p2, = ax3.plot(rharr[sampleloc, sampleloc, 1:frame_range],
color='blue',
label='output')
# p5, = ax3.plot(A1[sampleloc,sampleloc,1:frame_range], color='green', label='A1')
# p6, = ax3.plot(A2[sampleloc,sampleloc,1:frame_range], color='red', label='A2')
ax3.set_xlabel('frame number')
# ax1.set_ylabel('image intensity', color='black')
ax3.set_ylabel('response', color='blue')
lines = [p3, p4, p2]
labs = [l.get_label() for l in lines]
ax3.legend(lines, labs, loc='best')
if (plot):
plt.savefig(outputfilen + '_out.png')
else:
plt.show()
plt.clf()
def basic_stimulus(inputfilen, outputfilen, preferdir=True, plot=True):
frame_range = 240
with open(inputfilen, 'rb') as input:
st = pickle.load(input).movie
sw_nf = md.NormalFlow_x()
sw_nf.input = st
sw_nf.run()
print(sw_nf.result.shape)
sw_bs = md.MD_borst()
sw_bs.input = st
sw_bs.run()
print(sw_bs.result.shape)
sw2 = rh.RH(
inputframe=st[..., 0],
# A1_phase = 'cos',
# A2_phase = 'sin',
A1_sigma=5,
A2_sigma=5,
# A1_Lambda = 4.0,
# A2_Lambda = 4.0,
fl=0.25,
fh=0.25,
# winsize = 2
)
# sw2.print_gabor()
rharr = np.empty_like(st)
A2B1 = np.empty_like(st)
A1B2 = np.empty_like(st)
A1 = np.empty_like(st)
A2 = np.empty_like(st)
B1 = np.empty_like(st)
B2 = np.empty_like(st)
# sw2.print_gabor('gab3')
if (preferdir):
for i in range(0, frame_range):
rharr[..., i] = sw2.run(st[..., i])
A2B1[..., i] = sw2.A2B1
A1B2[..., i] = sw2.A1B2
A1[..., i] = sw2.convA1
A2[..., i] = sw2.convA2
B1[..., i] = sw2.lowpassA1
B2[..., i] = sw2.lowpassA2
print('frame', i)
else:
for i in range(0, frame_range):
rharr[..., i] = sw2.run(st[..., frame_range - i])
A2B1[..., i] = sw2.A2B1
A1B2[..., i] = sw2.A1B2
A1[..., i] = sw2.convA1
A2[..., i] = sw2.convA2
B1[..., i] = sw2.lowpassA1
B2[..., i] = sw2.lowpassA2
print('frame', i)
rharr = rharr * 1
A2B1 = A2B1 * 1
A1B2 = A1B2 * 1
SMALL_SIZE = 14
fig, host = plt.subplots()
fig.subplots_adjust(right=0.75)
par1 = host.twinx()
par2 = host.twinx()
par3 = host.twinx()
par2.spines["right"].set_position(("axes", 1.1))
par3.spines["right"].set_position(("axes", 1.2))
make_patch_spines_invisible(par2)
par2.spines["right"].set_visible(True)
sampleloc = 45
p1, = host.plot(st[sampleloc, sampleloc, 1:frame_range],
color='black',
label='input image')
# p2, = ax3.plot(rharr[sampleloc,sampleloc,1:frame_range], color='blue', label='st-Reichardt')
# p3, = ax3.plot(sw_bs.result[sampleloc+1,sampleloc+1,0:frame_range], color='green', label='Borst')
# p4, = ax4.plot(sw_nf.result[sampleloc,sampleloc,1:frame_range], color='red', label='LK-method')
p2, = par1.plot(rharr[sampleloc, sampleloc, 1:frame_range],
color='blue',
label='st-Reichardt')
p3, = par2.plot(sw_bs.result[sampleloc + 1, sampleloc + 1, 0:frame_range],
'--',
color='red',
label='Borst')
p4, = par3.plot(sw_nf.result[sampleloc, sampleloc, 1:frame_range],
'--',
color='green',
label='LK-method')
# host.set_xlim(0, 2)
host.set_ylim(-0.5, 2)
par1.set_ylim(-50, 200)
par2.set_ylim(-2, 8)
par3.set_ylim(-0.5, 2)
host.set_xlabel('frame number', size=SMALL_SIZE)
host.set_ylabel('image intensity', color='black', size=SMALL_SIZE)
par1.set_ylabel('st-Reichardt response', color='blue', size=SMALL_SIZE)
par2.set_ylabel('Borst response', color='red', size=SMALL_SIZE)
par3.set_ylabel('optical flow value', color='green', size=SMALL_SIZE)
host.yaxis.label.set_color(p1.get_color())
par1.yaxis.label.set_color(p2.get_color())
par2.yaxis.label.set_color(p3.get_color())
par3.yaxis.label.set_color(p4.get_color())
plt.rc('font', size=SMALL_SIZE) # controls default text sizes
tkw = dict(size=4, width=1.5)
host.tick_params(axis='y', colors=p1.get_color(), **tkw)
par1.tick_params(axis='y', colors=p2.get_color(), **tkw)
par2.tick_params(axis='y', colors=p3.get_color(), **tkw)
# par3.tick_params(axis='y', colors='red', **tkw)
host.tick_params(axis='x', **tkw)
fig.set_size_inches(15.5, 7.5)
lines = [p1, p2, p3, p4]
host.legend(lines, [l.get_label() for l in lines])
if (plot):
plt.savefig(outputfilen + '.png')
else:
plt.show()
plt.clf()