def read_ply(self, file_name):
num_samples = self.num_samples // len(self.files_list)
if self.file_index == len(self.files_list) - 1:
num_samples = num_samples + (self.num_samples - (num_samples * len(self.files_list)))
root, ext = os.path.splitext(file_name)
if not os.path.isfile(root + ".npy"):
ply = PlyData.read(file_name)
vertex = ply['vertex']
(x, y, z) = (vertex[t] for t in ('x', 'y', 'z'))
points = zip(x.ravel(), y.ravel(), z.ravel())
np.save(root + ".npy", points)
else:
points = np.load(root + ".npy")
#load normals
if os.path.isfile(root + "_normals" + ".ply"):
if not os.path.isfile(root + "_normals" + ".npy"):
ply1 = PlyData.read(root + "_normals" + ".ply")
vertex = ply1['vertex']
(nx, ny, nz) = (vertex[t] for t in ('nx', 'ny', 'nz'))
self.normals = np.asarray(zip(nx.ravel(), ny.ravel(), nz.ravel()))
np.save(root + "_normals" + ".npy", self.normals)
else:
self.normals = np.load(root + "_normals" + ".npy")
if self.add_noise:
self.data = utils.add_noise_normal(points, std=self.nois_std)
else:
self.data = np.asarray(points)
self.pc_diameter = utils.get_pc_diameter(self.data)
self.l = self.relL*self.pc_diameter
rot = utils.angle_axis_to_rotation(self.rotation_angle, self.rotation_axis)
self.data = utils.transform_pc(self.data, rot)
#plotutils.show_pc(self.data)
#mlab.show()
#TODO: better sampling
print "sampling file: ", file_name
self.samples, self.sample_indices = Sampler.sample(self.data, -1, min_num_point=-1, file_name=file_name, sampling_algorithm=self.sampling_algorithm)
#self.samples, self.sample_indices = Sampler.sample(self.data, -1, num_samples, file_name=file_name, sampling_algorithm=self.sampling_algorithm)
#self.samples = self.samples[0:num_samples]
#self.sample_indices = self.sample_indices[0:num_samples]
self.tree = spatial.KDTree(self.data)
return self.data
def readMesh_PLY(filename, output="soup"):
if output != "soup":
raise Exception("Mesh types other than soup not yet supported")
# Read the actual file
# TODO This takes a long time, maybe try to replace with something faster of my own?
plydata = PlyData.read(filename)
# Read vertices
# If the mesh has more than three columns of vertex data, ignore the later columns
# (for instance, Stanford Mesh Repo meshes store intensity and confidence here)
nVerts = plydata["vertex"].count
verts = np.zeros((nVerts, 3))
verts[:, 0] = np.array(plydata["vertex"].data["x"])
verts[:, 1] = np.array(plydata["vertex"].data["y"])
verts[:, 2] = np.array(plydata["vertex"].data["z"])
# Read faces
faces = make2d(plydata["face"].data["vertex_indices"])
# Build a mesh from these vertices and faces
mesh = TriSoupMesh(verts, faces)
return mesh
def main():
res_folder = opt.result_dir
ply_folder = opt.scan_path
output_dir = opt.output_dir
os.makedirs(output_dir, exist_ok=True)
reader_ins = Benchmark_reader(res_folder)
for folder in os.listdir(res_folder):
if os.path.isdir(os.path.join(res_folder, folder)):
continue
print(folder)
# ply reader
ply_file = os.path.join(ply_folder, folder.split('.')[0], folder.split('.')[0]+'_vh_clean_2.ply')
ply_data = PlyData.read(ply_file)
points = []
for point in ply_data.elements[0].data:
points.append([point[0], point[1], point[2]])
points = np.array(points)
colors = np.zeros_like(points)
# instance reader
instances = reader_ins[folder]
for instance_idx, instance_key in enumerate(instances.keys()):
r, g, b = create_color_palette()[int((instance_idx + 1)%41)]
colors[instances[instance_key]['points'].nonzero()[0].astype(np.int32)] = [r,g,b]
output_file = os.path.join(output_dir, folder.split('.')[0] + '.ply')
write_ply(points, colors, None, output_file)
def ply_plot(ply_file, opacity = 1, color = (1,1,1)):
ply = PlyData.read(ply_file)
'''
Plot vertices and triangles from a PlyData instance. Assumptions:
`ply' has a 'vertex' element with 'x', 'y', and 'z'
properties;
`ply' has a 'face' element with an integral list property
'vertex_indices', all of whose elements have length 3.
'''
vertex = ply['vertex'].data
(x, y, z) = (vertex[t] for t in ('x', 'y', 'z'))
# mlab.points3d(x, y, z, color=(1, 1, 1), mode='point')
tri_idx = ply['face'].data['vertex_indices']
idx_dtype = tri_idx[0].dtype
triangles = numpy.fromiter(tri_idx, [('data', idx_dtype, (3,))],
count=len(tri_idx))['data']
mesh = mlab.triangular_mesh(x, y, z, triangles,
color=color,
opacity = opacity)
return mesh
def main():
parser = ArgumentParser()
parser.add_argument("ply_filename")
args = parser.parse_args()
plot(PlyData.read(args.ply_filename))
def read_ply(self, file_name, num_samples=1000, sample_class_start=0, add_noise =False,
noise_prob=0.3, noise_factor=0.02, noise_std=0.1, sampling_algorithm=SampleAlgorithm.Uniform,
rotation_axis=[0, 0, 1], rotation_angle=0):
root, ext = os.path.splitext(file_name)
if not os.path.isfile(root + ".npy"):
ply = PlyData.read(file_name)
vertex = ply['vertex']
(x, y, z) = (vertex[t] for t in ('x', 'y', 'z'))
points = zip(x.ravel(), y.ravel(), z.ravel())
np.save(root + ".npy", points)
else:
points = np.load(root + ".npy")
#load normals
if os.path.isfile(root + "_normals" + ".ply"):
if not os.path.isfile(root + "_normals" + ".npy"):
ply1 = PlyData.read(root + "_normals" + ".ply")
vertex = ply1['vertex']
(nx, ny, nz) = (vertex[t] for t in ('nx', 'ny', 'nz'))
self.normals = np.asarray(zip(nx.ravel(), ny.ravel(), nz.ravel()))
np.save(root + "_normals" + ".npy", self.normals)
else:
self.normals = np.load(root + "_normals" + ".npy")
if add_noise:
print "adding noise to model.."
mr = utils.model_resolution(np.array(points))
#mr = 0.404
print "model resolution: ", mr
self.data = utils.add_noise_normal(np.array(points), mr, noise_std)
else:
self.data = np.asarray(points)
rot = utils.angle_axis_to_rotation(rotation_angle, rotation_axis)
self.data = utils.transform_pc(self.data, rot)
#plotutils.show_pc(self.data)
#mlab.show()
#TODO: better sampling
self.samples, self.sample_indices = Sampler.sample(self.data, -1, num_samples-1, file_name=file_name, pose=rot, sampling_algorithm=sampling_algorithm)
self.tree = spatial.KDTree(self.data)
self.sample_class_start = sample_class_start
self.sample_class_current = sample_class_start
self.num_samples = self.samples.shape[0]
print "num samples: ", self.num_samples
logging.basicConfig(filename='example.log',level=logging.DEBUG)
return self.data
def main():
parser = ArgumentParser()
parser.add_argument('ply_filename')
args = parser.parse_args()
plot(PlyData.read(args.ply_filename))
mlab.show()
def write_read(ply, tmpdir, name="test.ply"):
"""
Utility: serialize/deserialize a PlyData instance through a
temporary file.
"""
filename = tmpdir.join(name)
ply.write(str(filename))
return PlyData.read(str(filename))
def test_write_stream(tmpdir, tet_ply_txt):
ply0 = tet_ply_txt
test_file = tmpdir.join("test.ply")
with test_file.open("wb") as f:
tet_ply_txt.write(f)
ply1 = PlyData.read(str(test_file))
verify(ply0, ply1)
def read_str(string, tmpdir, name="test.ply"):
"""
Utility: create a PlyData instance from a string.
"""
filename = tmpdir.join(name)
with filename.open("wb") as f:
f.write(string)
return PlyData.read(str(filename))
def read_ply_cloud(filename):
ply_data = PlyData.read(filename)
points = ply_data['vertex'].data.copy()
cloud = np.empty([2048, 3])
for i in range(len(points)):
point = points[i]
p = np.array([point[0], point[1], point[2]])
cloud[i] = p
return np.array(cloud)
def read_mesh_vertices(filename):
assert os.path.isfile(filename)
with open(filename, 'rb') as f:
plydata = PlyData.read(f)
num_verts = plydata['vertex'].count
vertices = np.zeros(shape=[num_verts, 3], dtype=np.float32)
vertices[:,0] = plydata['vertex'].data['x']
vertices[:,1] = plydata['vertex'].data['y']
vertices[:,2] = plydata['vertex'].data['z']
return vertices
def load_ply(self, path):
plyData = PlyData.read(path)
data = plyData['vertex']
vertex_number = len(data[:])
vertex_data = np.zeros((vertex_number,3),dtype='int64')
for i in range(0, vertex_number):
vertex_data[i][0] = data[i][0]
vertex_data[i][1] = data[i][1]
vertex_data[i][2] = data[i][2]
#print(vertex_data)
return vertex_data
def main():
parser = ArgumentParser()
parser.add_argument('ply_filename')
args = parser.parse_args()
# file1 = open(args.ply_filename,'r')
# plot(PlyData.read())
# plot(file1)
plot(PlyData.read(args.ply_filename))
def test_obj_info(tmpdir):
ply0 = PlyData([], text=True, obj_info=["test obj_info"])
test_file = tmpdir.join("test.ply")
ply0.write(str(test_file))
ply0_str = test_file.read("rb").decode("ascii")
assert ply0_str.startswith("ply\r\nformat ascii 1.0\r\n" "obj_info test obj_info\r\n")
ply1 = PlyData.read(str(test_file))
assert len(ply1.obj_info) == 1
assert ply1.obj_info[0] == "test obj_info"
def convert(inputfp, outputfp):
print 'converting %s to %s'%(inputfp, outputfp)
plydata = PlyData.read(inputfp)
vertex = np.array([each for each in plydata.elements[0].data.tolist()], dtype=np.float32)
faces = plydata.elements[1].data.tolist()
faces = np.array([each[0] for each in faces], dtype=np.int32)
gv = gifti.GiftiDataArray.from_array(vertex, intent=1008)
gf = gifti.GiftiDataArray.from_array(faces, intent=1009)
g = gifti.GiftiImage()
g.add_gifti_data_array(gv)
g.add_gifti_data_array(gf)
gio.write(g, outputfp)
def ply_to_patch(ply_file_path,connection_string,pcid,writing_query,additional_offset,grouping_rules):
""" This function read a ply file, group points into 1M3 patches, convert patches
"""
from plyfile import PlyData
import datetime
print '\t working on ply file : ',ply_file_path
print '\t importing ply file ',datetime.datetime.now()
plydata = PlyData.read(ply_file_path)
print '\t grouping points',datetime.datetime.now()
numpy_spec_patch = grouping_ply_data(plydata, grouping_rules)
#to order the patch
#sorted_points = np.sort(patch[1], axis=0, kind='quicksort', order=('GPS_time'))
#send patch to database*
print '\t sending patch to database ply file ',datetime.datetime.now()
return making_pgpatch( numpy_spec_patch, connection_string, pcid, writing_query, ply_file_path, additional_offset)
def load_from_ply(self, filename, rotationFlag):
plydata = PlyData.read(filename)
self.plydata = plydata
self.f = np.vstack(plydata['face'].data['vertex_indices'])
if rotationFlag is not None:
x = -plydata['vertex'].data['z']
y = plydata['vertex'].data['x']
z = -plydata['vertex'].data['y']
else:
x = plydata['vertex'].data['x']
y = plydata['vertex'].data['y']
z = plydata['vertex'].data['z']
self.v = np.zeros([x.size, 3])
self.v[:,0] = x
self.v[:,1] = y
self.v[:,2] = z
def readPLY(fileName):
plydata = PlyData.read(fileName)
verts = []
faces = []
for ele in plydata.elements:
if ele.name == 'vertex':
for v in plydata['vertex']:
if len(v) == 3:
verts.append(np.asarray([v[0], v[1], v[2]]))
else:
verts.append(np.asarray([v[0][0], v[0][1], v[0][2]]))
if ele.name == 'face':
for f in plydata['face']:
if len(f) == 3:
faces.append(np.asarray([f[0], f[1], f[2]]))
else:
faces.append(np.asarray([f[0][0], f[0][1], f[0][2]]))
normals = []
return verts, faces,normals
def readPlyFile(filename):
"""
Usese plyfile python pacakage to read a ply file.
Gets around issues with pcl having a bad ply writer for pointclouds
:param filename:
:type filename: str
:return: vtkPolyData
:rtype:
"""
from plyfile import PlyData
plydata = PlyData.read(filename)
vertex_data = plydata['vertex'].data # numpy array with fields ['x', 'y', 'z']
pts = np.zeros([vertex_data.size, 3])
pts[:, 0] = vertex_data['x']
pts[:, 1] = vertex_data['y']
pts[:, 2] = vertex_data['z']
return vnp.numpyToPolyData(pts)
def run_point_cloud(self, voxel_world_id, threshold):
import voxel_globe.tools
import voxel_globe.meta.models as models
import boxm2_adaptor
import boxm2_mesh_adaptor
from plyfile import PlyData
voxel_world = models.VoxelWorld.objects.get(id=voxel_world_id)
with voxel_globe.tools.task_dir("voxel_viewer") as processing_dir:
scene_path = os.path.join(voxel_world.directory, "scene.xml")
scene, cache = boxm2_adaptor.load_cpp(scene_path)
ply_filename = os.path.join(processing_dir, "model.ply")
boxm2_mesh_adaptor.gen_color_point_cloud(scene, cache, ply_filename, 0.5, "")
ply = PlyData.read(str(ply_filename))
return ply.elements[0].data
def sample_ISS(file_name, min_num_point, pose):
root, ext = os.path.splitext(file_name)
in_file = root + ".ply"
out_file = root + "_iss.ply"
if (not os.path.isfile(out_file)):
print "file doen't exits.................."
args = ["./iss_detect", in_file, out_file]
popen = subprocess.Popen(args, stdout=subprocess.PIPE)
popen.wait()
output = popen.stdout.read()
print output
pc = np.load(root + '.npy')
tree = spatial.KDTree(pc)
ply = PlyData.read(out_file)
vertex = ply['vertex']
(x, y, z) = (vertex[t] for t in ('x', 'y', 'z'))
pc_iss = np.asarray(zip(x.ravel(), y.ravel(), z.ravel()))
indices = np.zeros((pc_iss.shape[0],))
for pt_i, samplept in enumerate(pc_iss):
_, index = tree.query(samplept, k=1)
indices[pt_i] = index
pc_iss = utils.transform_pc(pc_iss, pose)
#min_num_point = min(int(pc_iss.shape[0] / 10), 200)
if min_num_point < 0:
#min_num_point = min(int(pc_iss.shape[0] / 10), 200)
#min_num_point = min(int(pc_iss.shape[0] / 1), 300)
#min_num_point = min(int(pc_iss.shape[0] / 1), 500)
min_num_point = int(pc_iss.shape[0] / 1)
if min_num_point >= pc_iss.shape[0]:
return pc_iss, indices
sample_step = int(pc_iss.shape[0] / min_num_point)
pc_iss_samples, _ = Sampler.sample_uniform(pc_iss, sample_step)
indices_samples, _ = Sampler.sample_uniform(indices, sample_step)
assert(pc_iss_samples.shape[0] == indices_samples.shape[0])
print ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,pc_iss shape:", pc_iss_samples.shape
return pc_iss_samples, indices_samples
def Slab_SVM(file_name, v, sigma):
plydata = PlyData.read(open(file_name))
vertexes = plydata['vertex'][:]
x = np.zeros((len(vertexes),3))
for i in range(len(vertexes)):
x[i] = np.array([vertexes[i][0],vertexes[i][1],vertexes[i][2]])
file_name_points = len(x)
x_0 = plydata['vertex']['x']
x_1 = plydata['vertex']['y']
x_2 = plydata['vertex']['z']
#createPlot(x_0,x_1,x_2,file_name)
y = np.zeros(len(x))
x,ignore_x,y,ignore_y=train_test_split(x,y,train_size=0.01,random_state=8)
len_x = len(x)
risk = createHyperPlane(x,v,sigma)
print file_name,'points:',file_name_points,'subsample:',len_x,'v',v,'sigma',sigma,'risk',risk
def ply2obj(ply_path: str, obj_path: str):
ply = PlyData.read(ply_path)
with open(obj_path, "w") as f:
f.write("# OBJ file\n")
verteces = ply["vertex"]
for v in verteces:
p = [v["x"], v["y"], v["z"]]
if "red" in v and "green" in v and "blue" in v:
c = [v["red"] / 256, v["green"] / 256, v["blue"] / 256]
else:
c = [0, 0, 0]
a = p + c
f.write("v %.6f %.6f %.6f %.6f %.6f %.6f \n" % tuple(a))
for v in verteces:
if "nx" in v and "ny" in v and "nz" in v:
n = (v["nx"], v["ny"], v["nz"])
f.write("vn %.6f %.6f %.6f\n" % n)
for v in verteces:
if "s" in v and "t" in v:
t = (v["s"], v["t"])
f.write("vt %.6f %.6f\n" % t)
if "face" in ply:
for i in ply["face"]["vertex_index"]:
f.write("f")
for j in range(i.size):
# ii = [ i[j]+1 ]
ii = [i[j] + 1, i[j] + 1, i[j] + 1]
# f.write(" %d" % tuple(ii) )
f.write(" %d/%d/%d" % tuple(ii))
f.write("\n")
def get_offset(self, fp):
try:
plydata = PlyData.read(fp)
if plydata['vertex'].count == 0:
return
cor = np.vstack((plydata['vertex']['x'], plydata['vertex']['y'],
plydata['vertex']['z'])).transpose()
if self.x_offset is None:
self.x_offset = min(cor[:, 0])
self.y_offset = min(cor[:, 1])
self.z_offset = min(cor[:, 2])
else:
self.x_offset = min(self.x_offset, min(cor[:, 0]))
self.y_offset = min(self.y_offset, min(cor[:, 1]))
self.z_offset = min(self.z_offset, min(cor[:, 2]))
except: # noqa: E722
cor, f = ply_parser(fp)
for i in range(0, len(f)):
for j in range(0, len(f[i])):
f[i][j] = int(f[i][j])
del f[i][0]
for face_index in f:
face_cor = cor[face_index]
if self.x_offset is None:
self.x_offset = min(face_cor[:, 0])
self.y_offset = min(face_cor[:, 1])
self.z_offset = min(face_cor[:, 2])
else:
self.x_offset = min(self.x_offset, min(face_cor[:, 0]))
self.y_offset = min(self.y_offset, min(face_cor[:, 1]))
self.z_offset = min(self.z_offset, min(face_cor[:, 2]))
def readPlyCloud(fileName, device=None):
plydata = PlyData.read(fileName)
verts = plydata["vertex"]
if verts.count == 0:
print("Warning: empty file!")
zeros = [0] * verts.count
ones = [1] * verts.count
minusones = [-1] * verts.count
x = verts["x"]
y = verts["y"]
z = verts["z"]
try:
nx = verts["nx"]
ny = verts["ny"]
nz = verts["nz"]
except ValueError:
nx = zeros
ny = zeros
nz = ones
try:
r = verts["red"] / 255.0
g = verts["green"] / 255.0
b = verts["blue"] / 255.0
except ValueError:
try:
r = verts["r"] / 255.0
g = verts["g"] / 255.0
b = verts["b"] / 255.0
except ValueError:
r = ones
g = ones
b = ones
points = torch.tensor([x,y,z,nx,ny,nz,r,g,b], dtype=torch.float, device=device)
points = points.transpose(0,1)
points[:,3:6] = normalize(points[:, 3:6], 1)
return points
def ply_to_obj(ply_path, obj_path, texture_size=(1024, 1024)):
ply_path = Path(ply_path)
obj_path = Path(obj_path)
ply_copied_path = obj_path.parent / ply_path.name
is_same = ply_copied_path == ply_path
if not is_same:
shutil.copy(ply_path, ply_copied_path)
ply = PlyData.read(ply_path)
ply_texture = None
for c in ply.comments:
if 'TextureFile' in c:
ply_texture = c.split(' ')[-1]
if ply_texture is None:
template = _get_template('template_vertexcolor_to_texture.mlx')
out_texture_path = obj_path.with_suffix('').name + '_texture.png'
script = template.format(out_texture_path=out_texture_path)
run_meshlab_script(ply_copied_path,
obj_path,
script,
cd_dir=obj_path.parent)
else:
template = _get_template('template_ply_texture_to_obj.mlx')
script = template
ply_texture_name = ply_texture.split('.')[0]
out_texture_path = obj_path.parent / (ply_texture_name +
'_texture.png')
shutil.copy(ply_path.parent / ply_texture, out_texture_path)
Image.open(out_texture_path).resize(
texture_size, resample=PIL.Image.BILINEAR).save(out_texture_path)
run_meshlab_script(ply_path, obj_path, template)
add_texture_to_mtl(obj_path)
if not is_same:
ply_copied_path.unlink()
return
def process_deprecated(self):
LOG.info("Processing dataset...")
for p in range(len(self.raw_paths)):
path_cloud = self.raw_paths[p]
path_joints = (path_cloud[:len(path_cloud) - 3] + "json").replace(
"cloud", "joints")
LOG.info("Processing cloud {0} out of {1}".format(
p, len(self.raw_paths)))
LOG.info(path_cloud)
LOG.info(path_joints)
hands_ = self.read_joints_json(path_joints)
labels = hands_["left_hand"] + hands_["right_hand"]
#print(labels)
with open(self.raw_paths[p], 'rb') as f:
print(self.raw_paths[p])
cloud_ = PlyData.read(f)
graph_ = self._create_graph(cloud_, self.k, labels)
torch.save(
graph_,
os.path.join(self.processed_dir,
"unrealhands_k{0}_{1}.pt".format(self.k, p)))
def prune(self):
call(["docker", "run", "-it", "-v", "./data:/data", "deeplab"])
plydata = PlyData.read("data/depthmaps/merged.ply")
filename = "data/reconstruction.json"
file = open(filename, "r")
data = json.loads(file.read())
for f in data.files:
fdata = json.loads(open(f.name, "r").read())
pco = pyclipper.PyclipperOffset()
pco.AddPath(fdata.path)
elements = []
for element in plydata.elements:
dist = element.x * f.gps.nx + element.y * f.gps.ny + element.y * f.gps.nz
xm = element.x + (f.gps.x - element.x) / dist
ym = element.y + (f.gps.y - element.y) / dist
s = pco.Execute([xm, ym])
if len(s) > 0:
elements.append(element)
plydata.elements = elements
PlyData(plydata).write("pure.ply")
def map_fn(param):
save_folder = param[0]
prediction_path = param[1]
name = prediction_path.split('/')[-1][:-4]
object = name[:-5]
gt_path = '../data/ModelNet10_normalize/' + object + '/train/' + name + '.off'
gt_path = '../data/perfect_models_test/mesh_off_norm/'+ name + '.off'
save_ply_path = save_folder + '/' + name + ".ply"
hd_cmd = 'meshlabserver -i %s -i %s -o %s -s ../third_party/calculate_HD.mlx -om vq' % (
gt_path, prediction_path, save_ply_path)
if os.system(hd_cmd + "> /dev/null 2>&1"):
print "cannot calculate HD for file: %s" % (prediction_path)
return None
with open(save_ply_path, 'rb') as f:
plydata = PlyData.read(f)
dist = np.asarray(plydata['vertex']['quality'])
dist = dist[:19244]
print len(dist)
# if len(dist)!=30000:
# print name
# return None
return dist
def generate_annotation_single(path,obj_id,diameter):
obj = objs[obj_id]
file_name = str(obj_id)
while len(file_name)<2:
file_name = '0'+file_name
processed = {'class':obj_id,'diam':diameter,'obj_name':obj,'img_path':os.path.join(path,'data',file_name,'rgb')}
processed['depth_path']=os.path.join(path,'data',file_name,'depth')
processed['mask_path']=os.path.join(path,'data',file_name,'mask')
gts = yaml.load(open(os.path.join(path,'data',file_name,'gt.yml')),Loader=yaml.FullLoader)
obj_mesh = PlyData.read(os.path.join(path,'models',f'obj_{file_name}.ply'))
xyz = np.stack((obj_mesh['vertex']['x'],obj_mesh['vertex']['y'],obj_mesh['vertex']['z']))
corners = get_3D_corners(xyz)
kps = np.concatenate((xyz.mean(1).reshape(3,1),corners),1)#objcet center of preprocessed is not at zero
bar = Bar('Processing',max = len(gts))
processed['trainlist']=os.path.join(path,'data',file_name,'train.txt')
processed['testlist']=os.path.join(path,'data',file_name,'test.txt')
processed['gt_num']=len(gts)
for idx in gts:
annos = gts[idx]
keep = []
anno ={}
for item in annos:
if item['obj_id'] != obj_id:
continue
projs = calculate_projections(kps,np.array(item['cam_R_m2c']),np.array(item['cam_t_m2c']))
item['kps'] = list(projs.reshape(-1))
keep.append(item)
img_name = str(idx)
img_name = '0'*(4-len(img_name))+img_name
assert os.path.exists(os.path.join(path,'data',file_name,'rgb',f'{img_name}.png'))
anno['img_name'] = img_name
anno['gt'] = keep
processed[idx] = anno
bar.next()
bar.finish()
json.dump(processed,open(f'{obj}.json','w'))
def read_mesh_from_zip(zip_fname, fname_in_zip):
"""
Read ply file from zip. Currently does not support normals!
Parameters
----------
zip_fname : str
fname_in_zip : str
Returns
-------
np.array, np.array, np.array
"""
with zipfile.ZipFile(zip_fname, allowZip64=True) as z:
txt = z.open(fname_in_zip)
plydata = PlyData.read(txt)
vert = plydata['vertex'].data
vert = vert.view((np.float32, len(vert.dtype.names))).flatten()
ind = np.array(
plydata['face'].data['vertex_indices'].tolist()).flatten()
# TODO: support normals
# norm = plydata['normals'].data
# norm = vert.view((np.float32, len(vert.dtype.names))).flatten()
return [ind, vert, None]
def read_ply(filename):
"""convert from a ply file. include the label and the object number"""
#---read the ply file--------
plydata = PlyData.read(filename)
xyz = np.stack([plydata['vertex'][n] for n in ['x', 'y', 'z']], axis=1)
try:
rgb = np.stack(
[plydata['vertex'][n] for n in ['red', 'green', 'blue']],
axis=1).astype(np.uint8)
except ValueError:
rgb = np.stack([plydata['vertex'][n] for n in ['r', 'g', 'b']],
axis=1).astype(np.float32)
if np.max(rgb) > 1:
rgb = rgb
try:
object_indices = plydata['vertex']['object_index']
labels = plydata['vertex']['label']
return xyz, rgb, labels, object_indices
except ValueError:
try:
labels = plydata['vertex']['label']
return xyz, rgb, labels
except ValueError:
return xyz, rgb
def rdp_fun(basic_path, eps):
for i in range(0, 16):
file_name = basic_path + '/debug/cut-' + str(i) + '.ply'
print('read cut cloud file name = ', file_name)
plydata = PlyData.read(file_name)
vertex = plydata['vertex']
(x, y, z) = (vertex[t] for t in ('x', 'y', 'z'))
x = np.reshape(x, (x.size, 1))
y = np.reshape(y, (y.size, 1))
z = np.reshape(z, (z.size, 1))
data = np.concatenate((x, y, z), axis=1)
rdp_result = rdp(data, epsilon=eps)
result_file_name = basic_path + '/debug/rdp_result' + str(i) + '.txt'
np.savetxt(result_file_name, rdp_result, fmt='%.7f', delimiter=' ')
#dbscan.dbscan(rdp_result.t, 0.05, 500)
print(rdp_result.shape)
print(rdp_result)
#size_block = 1 * sampling_frequency
#window = np.ones(size_block)
#window /= sum(window)
#
#reference_velocity.x = np.convolve(reference_velocity.x, window, mode='same')
#reference_velocity.y = np.convolve(reference_velocity.y, window, mode='same')
#reference_velocity.z = np.convolve(reference_velocity.z, window, mode='same')
#
#odometry_velocity.x = np.convolve(odometry_velocity.x, window, mode='same')
#odometry_velocity.y = np.convolve(odometry_velocity.y, window, mode='same')
#odometry_velocity.z = np.convolve(odometry_velocity.z, window, mode='same')
########################
# Load Terrain DEM
########################
plydata = PlyData.read(open(esa_arl_dem_file))
vertex = plydata['vertex'].data
[px, py, pz] = (vertex[t] for t in ('x', 'y', 'z'))
# define grid.
npts=100
xi = np.linspace(min(px), max(px), npts)
yi = np.linspace(min(py), max(py), npts)
# grid the data.
zi = griddata(px, py, pz, xi, yi, interp='linear')
#################
## RE-SAMPLE ##
def read_pcd(self, file_path):
""" Read in a point cloud from to-be-specified file format into numpy array """
file_name, file_extension = os.path.splitext(file_path)
if file_extension == ".yml" or file_extension == ".yaml":
stream = io.open(filepath, "r")
data_loaded = yaml.safe_load(stream)
#~ data_numpy = numpy.zeros((len(data_loaded), len(data_loaded[0].keys())))
#~ for index in range(len(data_loaded)):
#~ data_numpy[index] = data_loaded[index]["X"], data_loaded[index]["Y"], data_loaded[index]["Z"]
self.points[:] = data_loaded[:]["X"], data_loaded[:][
"Y"], data_loaded[:]["Z"]
elif file_extension == ".mat":
# QUICK HACK
#~ self.points = scipy.io.loadmat(file_path)["pc"] # inner_sheet.mat
self.points = scipy.io.loadmat(file_path)["pci"]
if self.points.shape[1] > 3:
raise Exception(
"Currently only point information can be loaded within this format! Switch to .xyz or .ply format!"
)
elif file_extension == ".xyz" or file_extension == ".txt":
self.points = numpy.loadtxt(file_path)
# We have xyz files with additional normals and maybe curvature as columns
if self.points.shape[1] > 3 and self.points.shape[1] < 7:
self.normals = self.points[:, 3:6]
self.points = numpy.delete(self.points, [3, 4, 5], axis=1)
else:
pass
if self.points.shape[1] > 6:
self.normals = self.points[:, 3:6]
self.curvature = self.points[:, 6]
self.curvature = self.curvature.reshape(
(self.curvature.shape[0], 1))
self.points = numpy.delete(self.points, [3, 4, 5, 6], axis=1)
else:
pass
elif file_extension == ".ply":
with open(file_path, "rb") as f:
plydata = PlyData.read(f)
properties = plydata.elements[0].data.dtype.names
self.points = numpy.zeros((plydata.elements[0].data.size, 3))
self.points.T[0], self.points.T[1], self.points.T[
2] = plydata.elements[0].data["x"][:], plydata.elements[
0].data["y"][:], plydata.elements[0].data["z"][:]
# We may have more than just point information
if len(properties) > 3:
self.normals = numpy.zeros(
(plydata.elements[0].data.size, 3))
self.normals.T[0], self.normals.T[1], self.normals.T[
2] = plydata.elements[0].data[
"nx"][:], plydata.elements[0].data[
"ny"][:], plydata.elements[0].data["nz"][:]
else:
pass
# We may have additional curvature information. Meshlab saves this under "quality"
if len(properties) > 6:
self.curvature = plydata.elements[0].data["quality"]
self.curvature = self.curvature.reshape(
(self.curvature.shape[0], 1))
else:
pass
elif file_extension == ".asc" or ".csv":
with open(file_path) as f:
data = csv.reader(f, delimiter=" ")
point_list = []
normals_list = []
curvature_list = []
for row in data:
point_list.append(row[0:3])
if len(row) > 3:
normals_list.append(row[3:6])
else:
pass
if len(row) > 6:
curvature_list.append(row[6])
else:
pass
self.points = numpy.array(point_list, dtype=numpy.float64)
if normals_list:
self.normals = numpy.array(normals_list,
dtype=numpy.float64)
else:
pass
if curvature_list:
self.curvature = numpy.array(curvature_list,
dtype=numpy.float64)
else:
pass
elif file_extension == ".stl": # This extension might get cancelled and we use meshlab for big data
model_mesh = mesh.Mesh.from_file(file_path)
model_mesh.vectors # This will give a triplet of vectors for each vertex
self.normals = model_mesh.normals
# TODO Clear this process! Which format do we get from this? We need to delete duplicate points, because we dont care about triangle information!
ipdb.set_trace()
self.points = numpy.vstack(
(model_mesh.v0, model_mesh.v1, model_mesh.v2))
else:
raise Exception(
"File format not supported. Only input .xyz or .ply point clouds!"
)
if self.points is None == True:
raise Exception("Loaded file was empty")
return
f1 = h5py.File("/home/misumi/Desktop/Write_hdf5File/hdf5_data/data_testtest1.h5", 'w')
#f2 = h5py.File("/home/misumi/Desktop/Write_hdf5File/hdf5_data/data_test2.h5", 'w')
#f3 = h5py.File("/home/misumi/Desktop/Write_hdf5File/hdf5_data/data_test3.h5", 'w')
#f4 = h5py.File("/home/misumi/Desktop/Write_hdf5File/hdf5_data/data_test4.h5", 'w')
#f5 = h5py.File("/home/misumi/Desktop/Write_hdf5File/hdf5_data/data_test5.h5", 'w')
#f = h5py.File("/home/misumi/Desktop/pointnet-master_misumi/misuminet5_ply_hdf5_2048/data_testing0.h5", 'w')
a_data1 = np.zeros((len(filename1), 2048, 3))
#a_data2 = np.zeros((len(filename2), 2048, 3))
#a_data3 = np.zeros((len(filename3), 2048, 3))
#a_data4 = np.zeros((len(filename4), 2048, 3))
#a_data5 = np.zeros((len(filename5), 2048, 3))
#a_pid = np.zeros((len(filenames), 2048), dtype = np.uint8)
for i in range(0, len(filename1)):
plydata = PlyData.read("/home/misumi/Desktop/Write_hdf5File/data/" + filename1[i])
#piddata = [line.rstrip() for line in open("./points_label/" + filenames[i] + ".seg", 'r')]
for j in range(0, 2048):
a_data1[i, j] = [plydata['vertex']['x'][j], plydata['vertex']['y'][j], plydata['vertex']['z'][j]]
#a_pid[i,j] = piddata[j]
data = f1.create_dataset("data1", data = a_data1)
#for i in range(0, len(filename2)):
# plydata = PlyData.read("/home/misumi/Desktop/Write_hdf5File/data/" + filename2[i])
# #piddata = [line.rstrip() for line in open("./points_label/" + filenames[i] + ".seg", 'r')]
# for j in range(0, 2048):
# a_data2[i, j] = [plydata['vertex']['x'][j], plydata['vertex']['y'][j], plydata['vertex']['z'][j]]
# #a_pid[i,j] = piddata[j]
#
#data = f2.create_dataset("data2", data = a_data2)
def __init__(self, dataset, kpnum, kptype, is_train, resume=True):
# Prepare
self.root = opj('/home/penggao/data/sixd', dataset)
self.kpnum = kpnum
self.kptype = kptype
self.is_train = is_train
self.pklpath = opj(self.root, 'libs/benchmark.%s-%d-%s.pkl' %
('train' if is_train else 'test', self.kpnum, self.kptype))
self.seq_num = 15
self.cam = np.zeros((3, 3))
self.models = dict()
self.models_info = dict()
self.kps = dict()
self.frames = dict()
# Try to load from disk
if resume == True:
try:
self._load_from_disk()
print("[LOG] Load SIXD from pkl file success")
return
except Exception as e:
print("[ERROR]", str(e))
print("[ERROR] Load from pkl file failed. Load all anew")
else:
print("[LOG] Load SXID all anew")
# Load camera matrix
print("[LOG] Load camera matrix")
with open(os.path.join(self.root, 'camera.yml')) as f:
content = yaml.load(f)
self.cam = np.array([[content['fx'], 0, content['cx']],
[0, content['fy'], content['cy']],
[0, 0, 1]])
# Load models and keypoints
print("[LOG] Load models and keypoints")
MODEL_ROOT = os.path.join(self.root, 'models')
KP_ROOT = os.path.join(
self.root, 'kps', str(self.kpnum), self.kptype)
with open(os.path.join(MODEL_ROOT, 'models_info.yml')) as f:
content = yaml.load(f)
for key, val in tqdm(content.items()):
name = '%02d' % int(key) # e.g. '01'
self.models_info[name] = val
ply_path = os.path.join(MODEL_ROOT, 'obj_%s.ply' % name)
data = PlyData.read(ply_path)
self.models[name] = np.stack((np.array(data['vertex']['x']),
np.array(data['vertex']['y']),
np.array(data['vertex']['z'])), axis=1)
kp_path = os.path.join(KP_ROOT, 'obj_%s.ply' % name)
data = PlyData.read(kp_path)
self.kps[name] = np.stack((np.array(data['vertex']['x']),
np.array(data['vertex']['y']),
np.array(data['vertex']['z'])), axis=1)
# Load annotations
print("[LOG] Load annotations")
for seq in tqdm(['%02d' % i for i in range(1, self.seq_num+1)]):
frames = list()
seq_root = opj(
self.root, 'train' if self.is_train else 'test', str(seq))
imgdir = opj(seq_root, 'rgb')
with open(opj(seq_root, 'gt.yml')) as f:
content = yaml.load(f)
for key, val in content.items():
frame = dict()
frame['path'] = opj(imgdir, '%04d.png' % int(key))
frame['annots'] = list()
obj_ids = []
for v in val:
annot = dict()
rot = np.array(v['cam_R_m2c']).reshape(3, 3)
tran = np.array(v['cam_t_m2c']).reshape(3, 1)
annot['pose'] = np.concatenate((rot, tran), axis=1)
# x1 y1 w h => x1 y1 x2 y2
bbox = np.array(v['obj_bb'])
bbox[2] += bbox[0]
bbox[3] += bbox[1]
annot['bbox'] = bbox
annot['obj_id'] = v['obj_id']
annot['kps'] = self.project_vertices(
self.kps['%02d' % v['obj_id']], annot['pose'])
frame['annots'].append(annot)
obj_ids.append(v['obj_id'])
frame['obj_ids'] = obj_ids
frames.append(frame)
self.frames[seq] = frames
# Save to pickle path
try:
self._save_to_disk()
print("[LOG] Save benchmark to disk")
except Exception as e:
print("[ERROR]", str(e))
print("[ERROR] Save to disk failed")
def load_ply(fname_in):
try:
plydata = PlyData.read(fname_in)
except:
raise ValueError('Could not read file %s' % fname_in)
return plydata
parser = argparse.ArgumentParser(description='Modifies ply file to be used for rendering.\n Example: ./plyScale.py -f input.ply -o out.ply -r 210 -cutX 10.0')
parser.add_argument('-f','--inputFile', help='Input file name', required=True)
parser.add_argument('-o','--outputFile', help='Output file name', required=True)
parser.add_argument('--lx', help='Desired size of bounding box (X axis)', required=False, default="0")
parser.add_argument('--ly', help='Desired size of bounding box (Y axis)', required=False, default="0")
parser.add_argument('--lz', help='Desired size of bounding box (Z axis)', required=False, default="0")
parser.add_argument('-r','--order', help='Reorder axis. By default 012, to swap x and z use 210', required=False, default="012")
helpStringForCut = 'Remove all the faces which are above specified value for %s. The origin is in the center of mass. If axis reordering was applied, axis are in new coordinates.'
parser.add_argument('--cutX', help=helpStringForCut%('X'), required=False, default="none")
parser.add_argument('--cutY', help=helpStringForCut%('Y'), required=False, default="none")
parser.add_argument('--cutZ', help=helpStringForCut%('Z'), required=False, default="none")
args = vars(parser.parse_args())
desiredBox = [float(args['lx']), float(args['ly']), float(args['lz'])]
plydata = PlyData.read(args['inputFile'])
vertices = plydata['vertex'].data
# swap coords
order = args['order']
if (order != "012"):
idx = [int(order[i]) for i in range(0, len(order))]
assert(len(idx) == 3)
print "Swapping axis!"
for i in range(0, len(vertices)):
v = copy.deepcopy(vertices[i])
for dim in range(0, 3):
vertices[i][dim] = v[ idx[dim] ]
# Current box
def load_ply_normal(filename, point_num):
plydata = PlyData.read(filename)
pc = plydata['normal'].data[:point_num]
pc_array = np.array([[x, y, z] for x, y, z in pc])
return pc_array
def __init__(self, config):
self.config = config
self.pp = pprint.PrettyPrinter(indent=4)
self.pp.pprint(config)
rgb_image_sub = message_filters.Subscriber(
"/camera/rgb/image_color",
ImageSensor_msg,
queue_size=10000
)
rgb_info_sub = message_filters.Subscriber(
"/camera/rgb/camera_info",
CameraInfo
)
depth_info_sub = message_filters.Subscriber(
"/camera/depth/camera_info",
CameraInfo
)
depth_image_sub = message_filters.Subscriber(
"/camera/depth/image",
ImageSensor_msg,
queue_size=10000
)
self.depth_scale = 100.0
self.depth_image_cache = message_filters.Cache(depth_image_sub, 10000, allow_headerless=False)
self.cv_bridge = CvBridge()
self.output_path = config['label_gen']['output_path']
model_path = get_model_path(config['label_gen']['model_dir'], config['label_gen']['object_name'], model_type="upright")
self.initial_rotation = get_initial_rotation(config['label_gen']['object_label'])
# model_path = "/media/aditya/A69AFABA9AFA85D9/Datasets/YCB_Video_Dataset/models/004_sugar_box/textured.ply"
# self.output_path = "./bag_output/sugar_1"
# self.initial_rotation = tf.transformations.quaternion_from_euler(0, 0, -math.pi/2) #sugar_1
# self.output_path = "./bag_output/sugar_2"
# self.initial_rotation = tf.transformations.quaternion_from_euler(0, 0, -2.0 * math.pi/3) #sugar_2
# self.output_path = "./bag_output/sugar_3"
# self.initial_rotation = tf.transformations.quaternion_from_euler(0, 0, 2.0 * math.pi/3) #sugar_3
# model_path = "/media/aditya/A69AFABA9AFA85D9/Datasets/YCB_Video_Dataset/models/035_power_drill/textured_upright.ply"
# self.output_path = "./bag_output/drill_1"
# self.initial_rotation = tf.transformations.quaternion_from_euler(0, 0, math.pi) #drill_1
# self.output_path = "./bag_output/drill_2"
# self.initial_rotation = tf.transformations.quaternion_from_euler(0, 0, 2.5/3 * math.pi) #drill_2
# self.output_path = "./bag_output/drill_3"
# self.initial_rotation = tf.transformations.quaternion_from_euler(0, 0, -2.5/3 * math.pi) #drill_3
# model_path = "/media/aditya/A69AFABA9AFA85D9/Datasets/YCB_Video_Dataset/models/005_tomato_soup_can/textured.ply"
# self.output_path = "./bag_output/soup_1"
# self.initial_rotation = tf.transformations.quaternion_from_euler(0, 0, 0) #soup_1
# model_path = "/media/aditya/A69AFABA9AFA85D9/Datasets/YCB_Video_Dataset/models/006_mustard_bottle/textured.ply"
# self.output_path = "./bag_output/mustard_1"
# self.initial_rotation = tf.transformations.quaternion_from_euler(0, 0, 0) #mustard_1
# self.output_path = "./bag_output/mustard_2"
# self.initial_rotation = tf.transformations.quaternion_from_euler(0, 0, 2.5/3 * math.pi) #mustard_2
# self.output_path = "./bag_output/sugar_3"
# self.initial_rotation = tf.transformations.quaternion_from_euler(0, 0, 2.0 * math.pi/3) #sugar_3
mkdir_if_missing(self.output_path)
self.world_frame = "/base_footprint"
self.camera_frame = "/camera_rgb_optical_frame"
self.tf_listener = tf.TransformListener()
self.camera_pose = None
self.rgb_camera_instrinc_matrix = None
self.depth_camera_instrinc_matrix = None
self.counter = 0
self.MAX_COUNTER = 400
self.pub_filtered_cloud = \
rospy.Publisher(
"image_node/filtered_cloud",
PointCloud2,
queue_size=10
)
self.pub_pose_cloud = \
rospy.Publisher(
"image_node/pose_cloud",
PointCloud2,
queue_size=10
)
# Read Model
cloud = PlyData.read(model_path).elements[0].data
cloud = np.transpose(np.vstack((cloud['x'], cloud['y'], cloud['z'])))
cloud_pose = pcl.PointCloud()
cloud_pose.from_array(cloud)
sor = cloud_pose.make_voxel_grid_filter()
sor.set_leaf_size(0.015, 0.015, 0.015)
cloud_pose = sor.filter()
self.mesh_cloud = np.asarray(cloud_pose)
# Z point up
self.object_height = np.max(self.mesh_cloud[:,2]) - np.min(self.mesh_cloud[:,2])
self.cloud_filter_params = {
"xmin" : 0.1,
"xmax" : 0.6,
"ymin" : 0.0,
"ymax" : 1.7,
"zmin" : 0.75, #drill, sugar, mustard
# "zmin" : 0.76, #soup can
"object_height" : self.object_height,
"downsampling_leaf_size" : 0.015
}
print(self.cloud_filter_params)
print ("Num points after downsample and filter : {}".format(
self.mesh_cloud.shape[0]))
rgb_image_sub.registerCallback(self.image_callback)
rgb_info_sub.registerCallback(self.rgb_info_callback)
depth_info_sub.registerCallback(self.depth_info_callback)
def read_ply(path):
ply_data = PlyData.read(path).elements[0][0:]
return ply_data
import sys
import os
sys.path.insert(0, './src/core')
from numpy.random import uniform, seed
from matplotlib.mlab import griddata
import matplotlib.pyplot as plt
from mpl_toolkits.mplot3d import axes3d
from matplotlib import cm
from pylab import *
import numpy as np
from plyfile import PlyData, PlyElement
import scipy
matplotlib.style.use('classic') #in matplotlib >= 1.5.1
plydata = PlyData.read(open(os.path.expanduser(decos_dem_file)))
vertex = plydata['vertex'].data
[px, py, pz] = (vertex[t] for t in ('x', 'y', 'z'))
# define grid.
npts=500
xi = np.linspace(min(px), max(px), npts)
yi = np.linspace(min(py), max(py), npts)
# grid the data.
zi = griddata(px, py, pz, xi, yi, interp='linear')
############
### PLOT ###
def __init__(self, args, voxel_path=None, bbox_path=None, shared_values=None):
super().__init__(args)
# read initial voxels or learned sparse voxels
self.voxel_path = voxel_path if voxel_path is not None else args.voxel_path
self.bbox_path = bbox_path if bbox_path is not None else getattr(args, "initial_boundingbox", None)
assert (self.bbox_path is not None) or (self.voxel_path is not None), \
"at least initial bounding box or pretrained voxel files are required."
self.voxel_index = None
self.scene_scale = getattr(args, "scene_scale", 1.0)
if self.voxel_path is not None:
# read voxel file
assert os.path.exists(self.voxel_path), "voxel file must exist"
if Path(self.voxel_path).suffix == '.ply':
from plyfile import PlyData, PlyElement
plyvoxel = PlyData.read(self.voxel_path)
elements = [x.name for x in plyvoxel.elements]
assert 'vertex' in elements
plydata = plyvoxel['vertex']
fine_points = torch.from_numpy(
np.stack([plydata['x'], plydata['y'], plydata['z']]).astype('float32').T)
if 'face' in elements:
# read voxel meshes... automatically detect voxel size
faces = plyvoxel['face']['vertex_indices']
t = fine_points[faces[0].astype('int64')]
voxel_size = torch.abs(t[0] - t[1]).max()
# indexing voxel vertices
fine_points = torch.unique(fine_points, dim=0)
# vertex_ids, _ = discretize_points(fine_points, voxel_size)
# vertex_ids_offset = vertex_ids + 1
# # simple hashing
# vertex_ids = vertex_ids[:, 0] * 1000000 + vertex_ids[:, 1] * 1000 + vertex_ids[:, 2]
# vertex_ids_offset = vertex_ids_offset[:, 0] * 1000000 + vertex_ids_offset[:, 1] * 1000 + vertex_ids_offset[:, 2]
# vertex_ids = {k: True for k in vertex_ids.tolist()}
# vertex_inside = [v in vertex_ids for v in vertex_ids_offset.tolist()]
# # get voxel centers
# fine_points = fine_points[torch.tensor(vertex_inside)] + voxel_size * .5
# fine_points = fine_points + voxel_size * .5 --> use all corners as centers
else:
# voxel size must be provided
assert getattr(args, "voxel_size", None) is not None, "final voxel size is essential."
voxel_size = args.voxel_size
if 'quality' in elements:
self.voxel_index = torch.from_numpy(plydata['quality']).long()
else:
# supporting the old style .txt voxel points
fine_points = torch.from_numpy(np.loadtxt(self.voxel_path)[:, 3:].astype('float32'))
else:
# read bounding-box file
bbox = np.loadtxt(self.bbox_path)
voxel_size = bbox[-1] if getattr(args, "voxel_size", None) is None else args.voxel_size
fine_points = torch.from_numpy(bbox2voxels(bbox[:6], voxel_size))
half_voxel = voxel_size * .5
# transform from voxel centers to voxel corners (key/values)
fine_coords, _ = discretize_points(fine_points, half_voxel)
fine_keys0 = offset_points(fine_coords, 1.0).reshape(-1, 3)
fine_keys, fine_feats = torch.unique(fine_keys0, dim=0, sorted=True, return_inverse=True)
fine_feats = fine_feats.reshape(-1, 8)
num_keys = torch.scalar_tensor(fine_keys.size(0)).long()
# ray-marching step size
if getattr(args, "raymarching_stepsize_ratio", 0) > 0:
step_size = args.raymarching_stepsize_ratio * voxel_size
else:
step_size = args.raymarching_stepsize
# register parameters (will be saved to checkpoints)
self.register_buffer("points", fine_points) # voxel centers
self.register_buffer("keys", fine_keys.long()) # id used to find voxel corners/embeddings
self.register_buffer("feats", fine_feats.long()) # for each voxel, 8 voxel corner ids
self.register_buffer("num_keys", num_keys)
#self.register_buffer("points_labels", fine_points) # voxel centers
self.register_buffer("keep", fine_feats.new_ones(fine_feats.size(0)).long()) # whether the voxel will be pruned
self.register_buffer("voxel_size", torch.scalar_tensor(voxel_size))
self.register_buffer("step_size", torch.scalar_tensor(step_size))
self.register_buffer("max_hits", torch.scalar_tensor(args.max_hits))
logger.info("loaded {} voxel centers, {} voxel corners".format(fine_points.size(0), num_keys))
# set-up other hyperparameters and initialize running time caches
self.embed_dim = getattr(args, "voxel_embed_dim", None)
self.deterministic_step = getattr(args, "deterministic_step", False)
self.use_octree = getattr(args, "use_octree", False)
self.track_max_probs = getattr(args, "track_max_probs", False)
self._runtime_caches = {
"flatten_centers": None,
"flatten_children": None,
"max_voxel_probs": None
}
# sparse voxel embeddings
if shared_values is None and self.embed_dim > 0:
self.values = Embedding(num_keys, self.embed_dim, None)
else:
self.values = shared_values
def create_angle_data(ply_path, json_path, out_dir, convt):
print("Start processing " + os.path.basename(os.path.basename(ply_path)))
windowSize = 4
xyScale = 3
# fetch input files from dirs
jsonf, plyf, gImf, pImf = find_files(ply_path, json_path)
if jsonf == [] or plyf == [] or gImf == [] or pImf == []:
return
if not os.path.isdir(out_dir):
os.mkdir(out_dir)
# if json file in dist dir, it's been processed, we skip it.
json_basename = os.path.basename(jsonf)
json_dst = os.path.join(out_dir, json_basename)
if os.path.exists(json_dst):
return
# obtain scan position and offsets from metadata
metadata = terra_common.lower_keys(terra_common.load_json(jsonf))
center_position = get_position(metadata)
yOffset = get_offset_from_metadata(metadata)
plyFile = PlyData.read(plyf)
plyData = plyFile.elements[0]
# get relationship between ply files and png files, that means each point in the ply file
# should have a corresponding pixel in png files, both depth png and gray png
pImg = cv2.imread(pImf, -1)
gImg = cv2.imread(gImf, -1)
pHei, pWid = pImg.shape[:2]
gHei, gWid = gImg.shape[:2]
if pWid == gWid:
gPix = np.array(gImg).ravel()
gIndex = (np.where(gPix > 32))
tInd = gIndex[0]
else:
pPix = np.array(pImg)
pPix = pPix[:, 2:].ravel()
pIndex = (np.where(pPix != 0))
gPix = np.array(gImg).ravel()
gIndex = (np.where(gPix > 33))
tInd = np.intersect1d(gIndex[0], pIndex[0])
nonZeroSize = tInd.size
pointSize = plyData.count
# if point size do not match, return
if nonZeroSize != pointSize:
return
# Initial data structures
gIndexImage = np.zeros(gWid * gHei)
gIndexImage[tInd] = np.arange(1, pointSize + 1)
gIndexImage_ = np.reshape(gIndexImage, (-1, gWid))
angle_data = []
for i in range(0, PLOT_COL_NUM):
angle_data.append(np.zeros((1, 6)))
# move ROI in a window size to do the meta analysis
for i in np.arange(0 + windowSize * xyScale, gWid - windowSize * xyScale,
windowSize * xyScale * 2):
for j in np.arange(0 + windowSize, gHei - windowSize, windowSize * 2):
# fetch points in the ROI
plyIndices = gIndexImage_[j - windowSize:j + windowSize + 1,
i - windowSize * xyScale:i +
windowSize * xyScale + 1]
plyIndices = plyIndices.ravel()
plyIndices_ = np.where(plyIndices > 0)
localIndex = plyIndices[plyIndices_[0]].astype('int64')
if plyIndices_[0].size < 50:
continue
localP = plyData.data[localIndex - 1]
xCoord = np.mean(localP["x"])
yCoord = np.mean(localP["y"])
# get ROI column number(32 columns in total)
area_ind = get_area_index(xCoord, yCoord, yOffset,
center_position[0], convt) - 1
if area_ind < 0:
continue
# calculate angle data and store point position in localNormal
localNormal = calcAreaNormalSurface(localP)
if localNormal != []:
angle_data[area_ind] = np.append(angle_data[area_ind],
[localNormal],
axis=0)
file_ind = 0
# save output data in npy files
for save_data in angle_data:
file_ind = file_ind + 1
out_basename = str(file_ind) + '.npy'
out_file = os.path.join(out_dir, out_basename)
np.save(out_file, save_data)
shutil.copyfile(jsonf, json_dst)
return
def main(argv):
if (len(argv) < 2):
print('filename to npy/ply is not passed as argument. terminated.')
return
pathToFile = argv[1]
filename, file_extension = os.path.splitext(pathToFile)
folder = os.path.dirname(pathToFile)
filename = os.path.basename(pathToFile)
# for the moment supports npy and ply
if (file_extension == '.npy'):
pclTime = np.load(pathToFile)
pclTimeSize = np.shape(pclTime)
elif (file_extension == '.npz'):
pclTime = np.load(pathToFile)
pclTime = pclTime['pred']
pclTimeSize = np.shape(pclTime)
elif (file_extension == '.ply'):
ply = PlyData.read(pathToFile)
vertex = ply['vertex']
(x, y, z) = (vertex[t] for t in ('x', 'y', 'z'))
pclTime = np.column_stack((x, y, z))
else:
print('unsupported file format.')
return
if (len(np.shape(pclTime)) < 3):
pclTimeSize = [1, np.shape(pclTime)[0], np.shape(pclTime)[1]]
pclTime.resize(pclTimeSize)
for pcli in range(0, pclTimeSize[0]):
pcl = pclTime[pcli, :, :]
pcl = standardize_bbox(pcl, 2048)
pcl = pcl[:, [2, 0, 1]]
pcl[:, 0] *= -1
pcl[:, 2] += 0.0125
xml_segments = [xml_head]
for i in range(pcl.shape[0]):
color = colormap(pcl[i, 0] + 0.5, pcl[i, 1] + 0.5,
pcl[i, 2] + 0.5 - 0.0125)
xml_segments.append(
xml_ball_segment.format(pcl[i, 0], pcl[i, 1], pcl[i, 2],
*color))
xml_segments.append(xml_tail)
xml_content = str.join('', xml_segments)
xmlFile = ("%s/%s_%02d.xml" % (folder, filename, pcli))
with open(xmlFile, 'w') as f:
f.write(xml_content)
f.close()
exrFile = ("%s/%s_%02d.exr" % (folder, filename, pcli))
if (not os.path.exists(exrFile)):
print(['Running Mitsuba, writing to: ', xmlFile])
subprocess.run([PATH_TO_MITSUBA2, xmlFile])
else:
print('skipping rendering because the EXR file already exists')
png = ("%s/%s_%02d.jpg" % (folder, filename, pcli))
print(['Converting EXR to JPG...'])
ConvertEXRToJPG(exrFile, png)
from mpl_toolkits.mplot3d import Axes3D
from plyfile import PlyData, PlyElement
# Every 100 data samples, we save 1. If things run too
# slow, try increasing this number. If things run too fast,
# try decreasing it... =)
reduce_factor = 10
# Look pretty...
matplotlib.style.use('ggplot')
# Load up the scanned armadillo
plyfile = PlyData.read('Datasets/stanford_armadillo.ply')
armadillo = pd.DataFrame({
'x':plyfile['vertex']['z'][::reduce_factor],
'y':plyfile['vertex']['x'][::reduce_factor],
'z':plyfile['vertex']['y'][::reduce_factor]
})
def do_PCA(armadillo):
#
# TODO: Write code to import the libraries required for PCA.
# Then, train your PCA on the armadillo dataframe. Finally,
# drop one dimension (reduce it down to 2D) and project the
# armadillo down to the 2D principal component feature space.
#
def read_ply(filename):
""" read XYZ point cloud from filename PLY file """
plydata = PlyData.read(filename)
pc = plydata['vertex'].data
pc_array = np.array([[x, y, z] for x, y, z in pc])
return pc_array
def color_code_depth_img(ply_path, json_path, out_dir, convt):
windowSize = 4
xyScale = 3
jsonf, plyf, gImf, pImf = find_files(ply_path, json_path)
if jsonf == [] or plyf == [] or gImf == [] or pImf == []:
return
if not os.path.isdir(out_dir):
os.mkdir(out_dir)
metadata = terra_common.lower_keys(terra_common.load_json(jsonf))
yOffset = get_offset_from_metadata(metadata)
gIm = Image.open(gImf)
[gWid, gHei] = gIm.size
# color code gray image
codeImg = Image.new("RGB", gIm.size, "black")
pix = np.array(gIm).ravel()
gIndex = (np.where(pix > 32))
nonZeroSize = gIndex[0].size
plydata = PlyData.read(plyf)
pointSize = plydata.elements[0].count
if nonZeroSize != pointSize:
print('Point counts do not match.')
return
gIndexImage = np.zeros(gWid * gHei)
gIndexImage[gIndex[0]] = np.arange(1, pointSize + 1)
gIndexImage_ = np.reshape(gIndexImage, (-1, gWid))
angle_data = []
for i in range(0, 32):
angle_data.append(np.zeros((1, 6)))
icount = 0
# get top angle
for i in np.arange(0 + windowSize * xyScale, gWid - windowSize * xyScale,
windowSize * xyScale * 2):
icount = icount + 1
jcount = 0
for j in np.arange(0 + windowSize, gHei - windowSize, windowSize * 2):
jcount = jcount + 1
plyIndices = gIndexImage_[j - windowSize:j + windowSize + 1,
i - windowSize * xyScale:i +
windowSize * xyScale + 1]
plyIndices = plyIndices.ravel()
plyIndices_ = np.where(plyIndices > 0)
localIndex = plyIndices[plyIndices_[0]].astype('int64')
if plyIndices_[0].size < 100:
continue
localP = plydata.elements[0].data[localIndex - 1]
yCoord = np.mean(localP["y"])
area_ind = get_area_index(yCoord, yOffset, convt) - 1
localNormal = calcAreaNormalSurface(localP)
if localNormal != []:
angle_data[area_ind] = np.append(angle_data[area_ind],
[localNormal],
axis=0)
hist_data = np.zeros((32, 90))
pix_height = np.zeros(32)
disp_window = np.zeros(32)
min_z = np.zeros(32)
ind = 0
for meta_angle in angle_data:
if meta_angle.size < 10:
continue
pix_height[ind] = get_scanned_height(meta_angle)
leaf_angle = remove_soil_points(meta_angle)
hist_data[ind] = gen_angle_hist_from_raw(meta_angle)
disp_window[ind] = np.argmax(hist_data[ind])
min_z[ind] = np.amin(meta_angle[1:, 5]) + 55
ind = ind + 1
# color code
for i in np.arange(0 + windowSize * xyScale, gWid - windowSize * xyScale,
windowSize * xyScale * 2):
icount = icount + 1
jcount = 0
for j in np.arange(0 + windowSize, gHei - windowSize, windowSize * 2):
jcount = jcount + 1
plyIndices = gIndexImage_[j - windowSize:j + windowSize + 1,
i - windowSize * xyScale:i +
windowSize * xyScale + 1]
plyIndices = plyIndices.ravel()
plyIndices_ = np.where(plyIndices > 0)
localIndex = plyIndices[plyIndices_[0]].astype('int64')
if plyIndices_[0].size < 100:
continue
localP = plydata.elements[0].data[localIndex - 1]
localNormal = calcAreaNormalSurface(localP)
yCoord = np.mean(localP["y"])
area_ind = get_area_index(yCoord, yOffset, convt) - 1
if localNormal == []:
continue
#if localNormal[5] < min_z[area_ind]:
# continue
#if angle_in_range(disp_window[area_ind], localNormal):
rgb = normals_to_rgb_2(localNormal)
codeImg.paste(rgb,
(i - windowSize * xyScale, j - windowSize,
i + windowSize * xyScale + 1, j + windowSize + 1))
#save_points(localP, '/Users/nijiang/Desktop/normal_plot.png', 4)
img1 = Image.open(gImf)
img1 = img1.convert('RGB')
img3 = Image.blend(img1, codeImg, 0.5)
save_png_file = os.path.join(out_dir, os.path.basename(gImf))
img3.save(save_png_file)
file_ind = 0
for save_data in angle_data:
file_ind = file_ind + 1
out_basename = str(file_ind) + '.npy'
out_file = os.path.join(out_dir, out_basename)
np.save(out_file, save_data)
json_basename = os.path.basename(jsonf)
json_dst = os.path.join(out_dir, json_basename)
shutil.copyfile(jsonf, json_dst)
return
def load_ply(path):
ply = PlyData.read(path)
data = ply.elements[0].data
x, y, z = data['x'], data['y'], data['z']
model = np.stack([x, y, z], axis=-1)
return model
def format_converter(current_path, model_name):
model_file = current_path + model_name
print("Converting file format for 3D point cloud model {}...\n".format(model_name))
model_name_base = os.path.splitext(model_file)[0]
# load the model file
try:
with open(model_file, 'rb') as f:
plydata = PlyData.read(f)
num_vertex = plydata.elements[0].count
print("Ply data structure: \n")
print(plydata)
print("\n")
print("Number of 3D points in current model: {0} \n".format(num_vertex))
except:
print("Model file does not exist!")
sys.exit(0)
#Parse the ply format file and Extract the data
Data_array_ori = np.zeros((num_vertex, 3))
Data_array_ori[:,0] = plydata['vertex'].data['x']
Data_array_ori[:,1] = plydata['vertex'].data['y']
Data_array_ori[:,2] = plydata['vertex'].data['z']
#sort point cloud data based on Z values
Data_array = np.asarray(sorted(Data_array_ori, key = itemgetter(2), reverse = False))
'''
#accquire data range
min_x = Data_array[:, 0].min()
max_x = Data_array[:, 0].max()
min_y = Data_array[:, 1].min()
max_y = Data_array[:, 1].max()
min_z = Data_array[:, 2].min()
max_z = Data_array[:, 2].max()
range_data_x = max_x - min_x
range_data_y = max_y - min_y
range_data_z = max_z - min_z
print (range_data_x, range_data_y, range_data_z)
print(min_x,max_x)
print(min_y,max_y)
print(min_z,max_z)
'''
#Normalize data
#min_max_scaler = preprocessing.MinMaxScaler(feature_range = (0,1000000))
min_max_scaler = preprocessing.MinMaxScaler(feature_range = (0,10000))
point_normalized = min_max_scaler.fit_transform(Data_array)
#point_normalized_scale = [i * 1 for i in point_normalized]
# Pass xyz to Open3D.o3d.geometry.PointCloud
pcd = o3d.geometry.PointCloud()
pcd.points = o3d.utility.Vector3dVector(point_normalized)
o3d.visualization.draw_geometries([pcd])
#Save modelfilea as ascii format in xyz
'''
def load_ply_data(filename):
plydata = PlyData.read(filename)
pc = plydata['vertex'].data
pc_array = np.array([[x, y, z]
for x, y, z, _nx, _ny, _nz, _r, _g, _b, _a in pc])
return pc_array
def get_point_cloud(point_cloud_id, number_points=None, history=None):
from voxel_globe.meta import models
from vpgl_adaptor import convert_local_to_global_coordinates_array, create_lvcs
import os
import numpy as np
from plyfile import PlyData
point_cloud = models.PointCloud.objects.get(id=point_cloud_id).history(history)
lvcs = create_lvcs(point_cloud.origin[1], point_cloud.origin[0], point_cloud.origin[2], 'wgs84')
ply = PlyData.read(str(os.path.join(point_cloud.directory, 'error.ply')))
data = ply.elements[0].data
if number_points:
try:
import heapq
data = np.array(heapq.nlargest(number_points, ply.elements[0].data,
key=lambda x:x['prob']))
except IndexError: #not a correctly formated ply file. HACK A CODE!
#This is a hack-a-code for Tom's ply file
data = ply.elements[0].data.astype([('x', '<f4'), ('y', '<f4'),
('z', '<f4'), ('red', 'u1'), ('green', 'u1'), ('blue', 'u1'),
('prob', '<f4')])
import copy
blah = copy.deepcopy(data['y'])
data['y'] = data['z']
data['z'] = -blah
blah = copy.deepcopy(data['blue'])
data['blue'] = data['green']
data['green'] = blah
data['prob'] = abs(data['x'] - 10 - sum(data['x'])/len(data['x'])) \
+ abs(data['y'] + 30 - sum(data['y'])/len(data['y'])) \
+ abs(data['z'] - sum(data['z'])/len(data['z']))
data['prob'] = max(data['prob']) - data['prob']
data = np.array(heapq.nlargest(number_points, data,
key=lambda x:x['prob']))
print data['prob']
lla = convert_local_to_global_coordinates_array(lvcs, data['x'].tolist(), data['y'].tolist(), data['z'].tolist());
latitude = np.array(lla[0])
longitude = np.array(lla[1])
altitude = np.array(lla[2])
color = map(lambda r,b,g:'#%02x%02x%02x' % (r, g, b), data['red'], data['green'], data['blue'])
return_data = {"latitude": latitude, "longitude": longitude,
"altitude": altitude, "color": color}
try:
return_data['le'] = data['le']
except ValueError:
return_data['le'] = (-np.ones(len(latitude))).tolist()
try:
return_data['ce'] = data['ce']
except ValueError:
return_data['ce'] = (-np.ones(len(latitude))).tolist()
return return_data
from mpl_toolkits.mplot3d import Axes3D
from plyfile import PlyData, PlyElement
from sklearn.decomposition import PCA, RandomizedPCA
# Every 100 data samples, we save 1. If things run too
# slow, try increasing this number. If things run too fast,
# try decreasing it... =)
reduce_factor = 100
# Look pretty...
matplotlib.style.use('ggplot')
# Load up the scanned armadillo
plyfile = PlyData.read('Datasets/stanford_armadillo.ply')
armadillo = pd.DataFrame({
'x': plyfile['vertex']['z'][::reduce_factor],
'y': plyfile['vertex']['x'][::reduce_factor],
'z': plyfile['vertex']['y'][::reduce_factor]
})
def do_PCA(armadillo):
#
# TODO: Write code to import the libraries required for PCA.
# Then, train your PCA on the armadillo dataframe. Finally,
# drop one dimension (reduce it down to 2D) and project the
# armadillo down to the 2D principal component feature space.
#
# NOTE: Be sure to RETURN your projected armadillo!
import os
import glob
import numpy as np
from mayavi import mlab
from plyfile import PlyData, PlyElement
folder_path = 'objects/tools/ply_wrench'
saving_folder_path = 'objects/tools/pts_wrench'
saving_ext = '.pts'
if not os.path.exists(saving_folder_path):
os.makedirs(saving_folder_path)
file_paths = glob.glob(folder_path + '/' + '*.ply')
for (i, file_path) in enumerate(file_paths):
with open(file_path, 'rb') as f:
file_name = os.path.basename(file_path)
name = os.path.splitext(file_name)[0]
saving_file_path = saving_folder_path + '/' + name + saving_ext
plydata = PlyData.read(f)
xyz = (plydata['vertex']['x'], plydata['vertex']['y'],
plydata['vertex']['z'])
xyz = np.array(xyz).transpose()
np.savetxt(fname=saving_file_path, X=xyz, fmt='%f')
print(file_name + ' Done')
parser.add_argument('--no_line',
'-n',
action='store_true',
default=False,
help='Do not plot lines between vertices and normals')
parser.add_argument('input',
metavar='INPUT',
type=str,
nargs=1,
help='input file')
args = parser.parse_args()
plydata = PlyData.read(args.input[0])
normals = get_vertex_normals(plydata)
vertices = get_vertices(plydata)
if args.sample < 1:
sample_indices = np.random.choice(vertices.shape[0],
int(vertices.shape[0] * args.sample),
replace=False)
vertices_sampled = vertices[sample_indices, :]
increased = args.length * normals[sample_indices, :] + vertices_sampled
else:
vertices_sampled = vertices
increased = args.length * normals + vertices
# plt.draw()
# fig = plt.figure()
# ax = fig.gca(projection='3d')
# ax.plot_trisurf(x, y, np.array(v).flatten(), cmap=cm.jet, linewidth=0.2)
# ax.set_title('Actual')
# plt.draw()
# plt.show()
return a, D, indices_chosen
def main(x, y, z, ftype):
if (ftype == 'quad'):
a = QuadSurfFit(x, y, z)
elif (ftype == 'lin'):
a = LinSurfFit(x, y, z)
return a
if __name__ == "__main__":
plydata = PlyData.read('example-1.ply')
x = (plydata['vertex']['x'])
y = (plydata['vertex']['y'])
z = (plydata['vertex']['z'])
a = main(x, y, z, 'quad')
print 'final a = '
print a
f.seek(0)
head = "ply\nformat ascii 1.0\ncomment VCGLIB generated\nelement vertex " + '2048' + "\nproperty float x\nproperty float y\nproperty float z\nelement face " + str(
0) + "\nproperty list uchar int vertex_indices\nend_header\n"
f.write(head)
for line in lines[:]:
f.write(line + "\n")
f.close()
if __name__ == '__main__':
import numpy as np
import time
for file in os.listdir('../../data'):
file = os.path.join('../../data', file)
plydata = PlyData.read(file)
a_data = []
for i in range(plydata['vertex'].count):
line = [
plydata['vertex']['x'][i], plydata['vertex']['y'][i],
plydata['vertex']['z'][i]
]
a_data.append(line)
pc = np.array(a_data)
coords = pc[:, :3]
pts = tf.expand_dims(tf.constant(pc), axis=0)
# pts = tf.transpose(pts,perm=[0, 2, 1])
# pc_input = tf.placeholder(tf.float32, shape=(2, 2048, 3))
#
pts = np.tile(pts, reps=(1, 1, 1))
return all_neighbors
def get_neighbors(root_id, plydata):
neighbors = Set()
for face in plydata.elements[1]:
if root_id in face[0]:
for neighbor in face[0]:
if root_id != neighbor:
neighbors.add(neighbor)
return neighbors
def get_faces(root_id, plydata):
faces = []
for face in plydata.elements[1]:
if root_id in face[0]:
faces.append(face)
return faces
def get_all_vertexes(plydata):
vertexes = Set()
for face in plydata.elements[1]:
vertexes.add(face[0][0])
return vertexes
if __name__=='__main__':
plydata = PlyData.read('big.ply')
print get_neighbors_second_level(0, plydata)
