def segment_rg2sp_graphcut(slic, seg, centers, labels_fg_prob, path_model,
coef_shape, coef_pairwise=5, allow_obj_swap=True,
prob_label_trans=(0.1, 0.03),
dict_thresholds=RG2SP_THRESHOLDS, debug_export=''):
""" wrapper for region growing method with some debug exporting """
if os.path.splitext(path_model)[-1] == '.npz':
shape_model = np.load(path_model)
else:
shape_model = pickle.load(open(path_model, 'rb'))
dict_debug = dict() if os.path.isdir(debug_export) else None
slic_prob_fg = seg_rg.compute_segm_prob_fg(slic, seg, labels_fg_prob)
labels_gc = seg_rg.region_growing_shape_slic_graphcut(
slic, slic_prob_fg, centers, (shape_model['mix_model'], shape_model['cdfs']),
shape_model['name'], coef_shape, coef_pairwise, prob_label_trans,
optim_global=True, allow_obj_swap=allow_obj_swap,
dict_thresholds=dict_thresholds, nb_iter=250,
debug_history=dict_debug)
if dict_debug is not None:
nb_iter = len(dict_debug['energy'])
for i in range(nb_iter):
fig = tl_visu.figure_rg2sp_debug_complete(seg, slic, dict_debug, i)
fig.savefig(os.path.join(debug_export, 'iter_%03d' % i))
plt.close(fig)
segm_obj = labels_gc[slic]
return segm_obj, centers, None
def test_region_growing_graphcut(self, name='insitu7545'):
""" """
if not os.path.exists(PATH_PKL_MODEL):
self.test_shape_modeling()
# file_model = pickle.load(open(PATH_PKL_MODEL, 'r'))
npz_file = np.load(PATH_PKL_MODEL, allow_pickle=True)
file_model = dict(npz_file[npz_file.files[0]].tolist())
logging.info('loaded model: %r', file_model.keys())
list_mean_cdf = file_model['cdfs']
model = file_model['mix_model']
img, _ = load_image_2d(os.path.join(PATH_IMAGE, name + '.jpg'))
seg, _ = load_image_2d(os.path.join(PATH_SEGM, name + '.png'))
annot, _ = load_image_2d(os.path.join(PATH_ANNOT, name + '.png'))
centers = pd.read_csv(os.path.join(PATH_CENTRE, name + '.csv'), index_col=0).values
centers[:, [0, 1]] = centers[:, [1, 0]]
slic = segment_slic_img2d(img, sp_size=25, relative_compact=0.3)
slic_prob_fg = compute_segm_prob_fg(slic, seg, LABELS_FG_PROB)
dict_debug = {}
labels_gc = region_growing_shape_slic_graphcut(
slic,
slic_prob_fg,
centers, (model, list_mean_cdf),
'set_cdfs',
coef_shape=5.,
coef_pairwise=15.,
prob_label_trans=[0.1, 0.03],
optim_global=False,
nb_iter=65,
allow_obj_swap=False,
dict_thresholds=DEFAULT_RG2SP_THRESHOLDS,
debug_history=dict_debug
)
segm_obj = labels_gc[slic]
logging.info('show debug: %r', dict_debug.keys())
for i in np.linspace(0, len(dict_debug['criteria']) - 1, 5):
fig = figure_rg2sp_debug_complete(seg, slic, dict_debug, int(i), max_size=5)
fig_name = 'RG2Sp_graph-cut_%s_debug-%03d.pdf' % (name, i)
fig.savefig(os.path.join(PATH_OUTPUT, fig_name), bbox_inches='tight', pad_inches=0)
plt.close(fig)
_ = adjusted_rand_score(annot.ravel(), segm_obj.ravel())
# self.assertGreaterEqual(score, 0.5)
expert_segm(name, img, seg, segm_obj, annot, str_type='RG2Sp_graph-cut')
def test_region_growing_greedy(self, name='insitu7545'):
""" """
if not os.path.exists(PATH_PKL_MODEL):
self.test_shape_modeling()
# file_model = pickle.load(open(PATH_PKL_MODEL, 'r'))
npz_file = np.load(PATH_PKL_MODEL, allow_pickle=True)
file_model = dict(npz_file[npz_file.files[0]].tolist())
logging.info('loaded model: %r', file_model.keys())
list_mean_cdf = file_model['cdfs']
model = file_model['mix_model']
img, seg, slic, centers, annot = load_inputs(name)
dict_debug = {}
slic_prob_fg = compute_segm_prob_fg(slic, seg, LABELS_FG_PROB)
labels_greedy = region_growing_shape_slic_greedy(
slic,
slic_prob_fg,
centers, (model, list_mean_cdf),
'set_cdfs',
coef_shape=5.,
coef_pairwise=15.,
prob_label_trans=[0.1, 0.03],
greedy_tol=3e-1,
allow_obj_swap=True,
dict_thresholds=DEFAULT_RG2SP_THRESHOLDS,
nb_iter=250,
debug_history=dict_debug)
segm_obj = labels_greedy[slic]
logging.info('show debug: %r', dict_debug.keys())
fig_size = (FIG_SIZE * np.array(img.shape[:2]) / np.max(img.shape))
for i in np.linspace(0, len(dict_debug['criteria']) - 1, 5):
fig, ax = plt.subplots(figsize=fig_size[::-1])
draw_rg2sp_results(ax, seg, slic, dict_debug, int(i))
fig_name = 'RG2Sp_greedy_%s_debug-%03d.pdf' % (name, i)
fig.savefig(os.path.join(PATH_OUTPUT, fig_name),
bbox_inches='tight',
pad_inches=0)
plt.close(fig)
score = adjusted_rand_score(annot.ravel(), segm_obj.ravel())
self.assertGreaterEqual(score, 0.5)
expert_segm(name, img, seg, segm_obj, annot, str_type='RG2Sp_greedy')
