target_segmentation_image = np.array(Image.open(target_segmentation_path))
index = 1
std_list = np.linspace(0, 0.02, size)
betas_list = np.linspace(0, 5000, size)
error_values = []
with open(seeds_file_path, 'rb') as seeds_file:
seeds = pickle.load(seeds_file)
for noise_std in std_list:
segmentation = Segmentation(image_array, 1, seeds, image_name)
segmentation.add_noise(0, noise_std)
segmentation.set_reference_segmentation(target_segmentation_path)
for beta in betas_list:
print(f'index = {index}, beta = {beta}, std = {noise_std}')
segmentation.beta = beta
segmentation.solve()
segmentation.build_segmentation_image()
error_values.append(segmentation.compute_error())
index += 1
X, Y = np.meshgrid(betas_list, std_list)
triang = mtri.Triangulation(np.reshape(
X, [size*size]), np.reshape(Y, [size*size]))
fig = plt.figure()
ax = plt.axes(projection='3d')
ax.set_xlabel(r'$\beta$')
ax.set_ylabel('noise')
ax.set_zlabel('error')
ax.set_zlim3d(0, 0.085)
ax.set_zticks([0.1])
