def plot_labelled_volume(self, label_map):
volume = self.tsdf_volume.get_volume()
grid_labelled = np.argmin(volume, axis=-1)
grid_occupied = (np.min(volume, axis=-1) < 0)
occupied_idxs = np.column_stack(
np.where(grid_occupied == True)) * self.resolution
occupied_labels = grid_labelled[np.where(grid_occupied == True)]
# voxel colors
colors = []
for label in occupied_labels:
colors.append(label_map[label])
colors = np.asarray(colors)
# voxel alphas
alphas = len(colors) * [255]
dataframe = dict()
dataframe['x'] = occupied_idxs[:, 0]
dataframe['y'] = occupied_idxs[:, 1]
dataframe['z'] = occupied_idxs[:, 2]
dataframe['red'] = colors[:, 0]
dataframe['green'] = colors[:, 1]
dataframe['blue'] = colors[:, 2]
dataframe['alpha'] = alphas
pcl = pd.DataFrame(data=dataframe, columns=dataframe.keys())
pcl = PyntCloud(pcl)
pcl.plot()
def reconstruction(
from_path=None,
to_path=None,
filtre=None,
k=None,
z_max=None,
r=None,
SW_TOF0=False,
SW_TOF1=False,
SW_TOF2=False,
internal=True,
distance=0,
start_circle=0,
stop_circle=None,
seuil=200,
speed_sample=48,
T=7617.43,
):
data_calculated = dispose_data(SW_TOF0=SW_TOF0,
SW_TOF1=SW_TOF1,
SW_TOF2=SW_TOF2,
internal=internal,
distance=distance,
start_circle=start_circle,
stop_circle=stop_circle,
path=from_path,
seuil=seuil,
speed_sample=speed_sample,
T=T)
point = pd.DataFrame(data_calculated, columns=[
'x',
'y',
'z',
])
cloud = PyntCloud(point)
if filtre is None:
print("Do not use any filter")
pass
elif filtre is "SOR":
cloud.apply_filter(filtre_SOR(point, k=k, z_max=z_max))
print("Use filter SOR")
elif filtre is "ROR":
cloud.apply_filter(filtre_ROR(point, k=k, r=r))
print("Use filter ROR")
cloud.to_file(to_path)
cloud.plot(return_scene=True)
print("Generated successfully, the PLY file has been stored in: ", to_path)
import numpy as np
import pandas as pd
from pyntcloud import PyntCloud
positions = np.random.rand(1000, 3) * 10
positions -= positions.mean(0)
points = pd.DataFrame(positions.astype(np.float32), columns=['x', 'y', 'z'])
points["red"] = (np.random.rand(1000) * 255).astype(np.uint8)
points["green"] = (np.random.rand(1000) * 255).astype(np.uint8)
points["blue"] = (np.random.rand(1000) * 255).astype(np.uint8)
cloud = PyntCloud(points)
lines = [{
"color": "red",
"vertices": [[0, 0, 0], [10, 0, 0]]
}, {
"color": "green",
"vertices": [[0, 0, 0], [0, 10, 0]]
}, {
"color": "blue",
"vertices": [[0, 0, 0], [0, 0, 10]]
}]
cloud.plot(polylines=lines)
#plt.show()
#point cloud
colourImg = imgL
indicesArray = np.moveaxis(np.indices((hl, wl)), 0, 2)
imageArray = np.dstack((indicesArray, colourImg)).reshape((-1, 5))
df = pd.DataFrame(imageArray, columns=["x", "y", "red", "green", "blue"])
depthImg = filteredImg
depthArray = np.array(depthImg[::-2])
df.insert(loc=2, column='z', value=depthArray)
df[['x', 'y', 'z']] = df[['x', 'y', 'z']].astype(float)
df[['red', 'green', 'blue']] = df[['red', 'green', 'blue']].astype(np.uint)
df['z'] = df['z'] * 0.5
cloud = PyntCloud(df)
cloud.plot()
#cv2.imshow("as", filt_Color)
#plt.imshow(image)
#plt.colorbar()
#plt.show()
cv2.waitKey()
cv2.destroyAllWindows()
main()
log
"""
fname = fname.replace('\\', '/').rsplit('/')[-1]
fname = fname[:fname.find('.')]
point_cloud_pynt.points.loc[feature_point_idxs, :].to_csv(
"output/{}_{}_{}_{}_{}.txt".format(fname, srm, nmrm, gamma_21,
gamma_32),
header=False,
index=False)
# with open("output/stats.txt", "a") as f:
# l = [class_, srm, nmrm, gamma_21, gamma_32, len(feature_point_idxs)]
# l = [str(e) for e in l]
# f.write(','.join(l)+'\n')
print(fname, nmrm, gamma_21, gamma_32, "There are",
len(feature_point_idxs), "feature points")
#%%
"""
Visualization
"""
point_cloud_pynt.plot()
# scene = point_cloud_pynt.plot(initial_point_size=0.01, return_scene=True)
# point_cloud_pynt_feature_points = PyntCloud(point_cloud_pynt.points.iloc[feature_point_idxs,:])
# scene.children += [point_cloud_pynt_feature_points]
# point_cloud_pynt_feature_points.plot(initial_point_size=0.5, scene = scene)
point_cloud_o3d = point_cloud_pynt.to_instance("open3d", mesh=False)
o3d.visualization.draw_geometries([point_cloud_o3d])
from pyntcloud import PyntCloud
import pandas as pd
df= pd.read_csv("../Data/V_20180816_I285_EB_run1(0).csv")
df.rename(columns={'X':'x',
'Y':'y',
'Z':'z'},
inplace=True)
anky_cloud = PyntCloud(df)
anky_cloud.plot()
def plot_points_from_torch(tensor):
cloud = PyntCloud(pd.DataFrame(tensor, columns=['x', 'y', 'z']))
cloud.plot(background='white', initial_point_size=0.003)
