1、主要参考
（1）官方稳定地址
Point cloud — Open3D 0.16.0 documentation
2、相关功能
2.1凸包（Convex hull）
（1）函数
compute_vertex_normals
create_from_triangle_mesh
（2）功能说明
点云的凸包是包含所有点的最小凸集。Open3D包含compute_convex_hull方法，用于计算点云的凸包。该实现基于Qhull。
（3）测试代码
在下面的示例代码中，我们首先从网格中采样一个点云，并计算作为三角形网格返回的凸包。然后，我们将凸包可视化为一个红色的LineSet。
1）本次测试还是那只兔子，原始的3d显示如下
import open3d as o3d
import numpy as np
#网络下载，官方例子
# bunny = o3d.data.BunnyMesh()
# mesh = o3d.io.read_triangle_mesh(bunny.path)
#或者下载后本地读取
#https://github.com/isl-org/open3d_downloads/releases/download/20220201-data/BunnyMesh.ply
plypath = "D:/RGBD_CAMERA/python_3d_process/BunnyMesh/BunnyMesh.ply"
# mesh = o3d.io.read_point_cloud(plypath) # path为文件路径 #直接读取点云的方法
mesh = o3d.io.read_triangle_mesh(plypath) # path为文件路径
o3d.visualization.draw_geometries([mesh])兔子的图，网络问题暂时上传不了
注意：读取时使用了read_triangle_mesh（三角网格）而不是read_point_cloud
2）方法一直接对兔子做凸包
import open3d as o3d
import numpy as np
#网络下载，官方例子
# bunny = o3d.data.BunnyMesh()
# mesh = o3d.io.read_triangle_mesh(bunny.path)
#或者下载后本地读取
#https://github.com/isl-org/open3d_downloads/releases/download/20220201-data/BunnyMesh.ply
plypath = "D:/RGBD_CAMERA/python_3d_process/BunnyMesh/BunnyMesh.ply"
# mesh = o3d.io.read_point_cloud(plypath) # path为文件路径 #直接读取点云的方法
mesh = o3d.io.read_triangle_mesh(plypath) # path为文件路径
# o3d.visualization.draw_geometries([mesh])
#(方法一)直接对兔子做凸包
hull, _ = mesh.compute_convex_hull()
hull_ls = o3d.geometry.LineSet.create_from_triangle_mesh(hull)
hull_ls.paint_uniform_color((1, 0, 0))
o3d.visualization.draw_geometries([mesh,hull_ls])
#(方法二)对兔子采样后做凸包
# pcl = mesh.sample_points_poisson_disk(number_of_points=2000)
# hull, _ = pcl.compute_convex_hull()
# hull_ls = o3d.geometry.LineSet.create_from_triangle_mesh(hull)
# hull_ls.paint_uniform_color((1, 0, 0))
# o3d.visualization.draw_geometries([pcl, hull_ls])
兔子的图如下
3）方法二，对兔子进行下采样后做凸包
import open3d as o3d
import numpy as np
#网络下载，官方例子
# bunny = o3d.data.BunnyMesh()
# mesh = o3d.io.read_triangle_mesh(bunny.path)
#或者下载后本地读取
#https://github.com/isl-org/open3d_downloads/releases/download/20220201-data/BunnyMesh.ply
plypath = "D:/RGBD_CAMERA/python_3d_process/BunnyMesh/BunnyMesh.ply"
# mesh = o3d.io.read_point_cloud(plypath) # path为文件路径 #直接读取点云的方法
mesh = o3d.io.read_triangle_mesh(plypath) # path为文件路径
# o3d.visualization.draw_geometries([mesh])
#(方法一)直接对兔子做凸包
# hull, _ = mesh.compute_convex_hull()
# hull_ls = o3d.geometry.LineSet.create_from_triangle_mesh(hull)
# hull_ls.paint_uniform_color((1, 0, 0))
# o3d.visualization.draw_geometries([mesh,hull_ls])
#(方法二)对兔子采样后做凸包
pcl = mesh.sample_points_poisson_disk(number_of_points=2000)
hull, _ = pcl.compute_convex_hull()
hull_ls = o3d.geometry.LineSet.create_from_triangle_mesh(hull)
hull_ls.paint_uniform_color((1, 0, 0))
o3d.visualization.draw_geometries([pcl, hull_ls])
兔子的图如下
（4）函数说明
sample_points_poisson_disk -- 泊松磁盘采样，下采样的一个方法
2.2 DBSCAN 分类集群（DBSCAN clustering）
（1）函数
cluster_dbscan(eps=0.02, min_points=10, print_progress=True)
（2）说明
给定一个点云，例如一个深度传感器，我们希望将局部点云数据集群分组在一起。为此，我们可以使用聚类算法。Open3D实现了DBSCAN [Ester1996]，这是一种基于密度的聚类算法。该算法在cluster_dbscan中实现，需要两个参数:eps定义到集群中邻居的距离，min_points定义形成集群所需的最小点数。函数返回标签，其中标签-1表示噪声。
（3）测试代码
import open3d as o3d
import numpy as np
from matplotlib import pyplot as plt
# 初次使用下载模型
# ply_point_cloud = o3d.data.PLYPointCloud()
# pcd = o3d.io.read_point_cloud(ply_point_cloud.path)
# 或者直接使用本地
plypath = "D:/RGBD_CAMERA/python_3d_process/DemoCropPointCloud/fragment.ply"
pcd = o3d.io.read_point_cloud(plypath) # path为文件路径
with o3d.utility.VerbosityContextManager(
o3d.utility.VerbosityLevel.Debug) as cm:
labels = np.array(
pcd.cluster_dbscan(eps=0.02, min_points=10, print_progress=True))
max_label = labels.max()
print(f"point cloud has {max_label + 1} clusters")
colors = plt.get_cmap("tab20")(labels / (max_label if max_label > 0 else 1))
colors[labels < 0] = 0
pcd.colors = o3d.utility.Vector3dVector(colors[:, :3])
o3d.visualization.draw_geometries([pcd],
zoom=0.455,
front=[-0.4999, -0.1659, -0.8499],
lookat=[2.1813, 2.0619, 2.0999],
up=[0.1204, -0.9852, 0.1215])
（4）测试结果
（稍微大一点的截图就没法上传，只能上小图）
该算法预先计算了所有点在半径内的所有邻点。如果所选eps太大，这可能需要大量内存。
（5）自己的例子，未剔除点云的处理结果
import open3d as o3d
import numpy as np
from matplotlib import pyplot as plt
# 初次使用下载模型
# ply_point_cloud = o3d.data.PLYPointCloud()
# pcd = o3d.io.read_point_cloud(ply_point_cloud.path)
# 或者直接使用本地
# plypath = "D:/RGBD_CAMERA/python_3d_process/DemoCropPointCloud/fragment.ply"
plypath = "D:/RGBD_CAMERA/python_3d_process/1_hezi.pcd"
pcd = o3d.io.read_point_cloud(plypath) # path为文件路径
pcd = pcd.remove_non_finite_points(True, False)#剔除无效值
with o3d.utility.VerbosityContextManager(
o3d.utility.VerbosityLevel.Debug) as cm:
labels = np.array(
# pcd.cluster_dbscan(eps=0.02, min_points=10, print_progress=True))
pcd.cluster_dbscan(eps=10, min_points=200, print_progress=True))
max_label = labels.max()
print(f"point cloud has {max_label + 1} clusters")
colors = plt.get_cmap("tab20")(labels / (max_label if max_label > 0 else 1))
colors[labels < 0] = 0
pcd.colors = o3d.utility.Vector3dVector(colors[:, :3])
o3d.visualization.draw_geometries([pcd],
zoom=0.455,
front=[-0.4999, -0.1659, -0.8499],
lookat=[2.1813, 2.0619, 2.0999],
up=[0.1204, -0.9852, 0.1215])
注意：其中标签-1表示噪声。用黑色表示
colors[labels < 0] = 0
2.3 平面分割（Plane segmentation）
（1）函数
segment_plane(distance_threshold=0.01,ransac_n=3,num_iterations=1000)
（2）功能说明
Open3D还支持使用RANSAC从点云中分割几何原始特征。为了找到点云中支持度最大的平面，我们可以使用segment_plane函数。该方法有三个参数:distance_threshold定义点到估计平面的最大距离，ransac_n定义估计平面时随机抽样的点的数量，num_iterations定义随机平面抽样和验证的频率。然后函数返回平面为(a,b,c,d)，这样对于平面上的每个点(x,y,z)我们得到ax+by+cz+d=0。该函数进一步返回索引列表。
（3）测试代码
import open3d as o3d
import numpy as np
from matplotlib import pyplot as plt
# 初次使用下载模型
pcd_point_cloud = o3d.data.PCDPointCloud()
pcd = o3d.io.read_point_cloud(pcd_point_cloud.path)
# 或者直接使用本地
# plypath = "D:/RGBD_CAMERA/python_3d_process/PCDPointCloud/fragment.pcd"
# pcd = o3d.io.read_point_cloud(plypath) # path为文件路径
plane_model, inliers = pcd.segment_plane(distance_threshold=0.01,
ransac_n=3,
num_iterations=1000)
[a, b, c, d] = plane_model
print(f"Plane equation: {a:.2f}x + {b:.2f}y + {c:.2f}z + {d:.2f} = 0")
inlier_cloud = pcd.select_by_index(inliers)
inlier_cloud.paint_uniform_color([1.0, 0, 0])
outlier_cloud = pcd.select_by_index(inliers, invert=True)
o3d.visualization.draw_geometries([inlier_cloud, outlier_cloud],
zoom=0.8,
front=[-0.4999, -0.1659, -0.8499],
lookat=[2.1813, 2.0619, 2.0999],
up=[0.1204, -0.9852, 0.1215])（4）测试结果
注意：其中标注为红色的是分割出来的平面
（5）自己的某个简单例子的测试
import open3d as o3d
import numpy as np
from matplotlib import pyplot as plt
# # 初次使用下载模型
# pcd_point_cloud = o3d.data.PCDPointCloud()
# pcd = o3d.io.read_point_cloud(pcd_point_cloud.path)
# # 或者直接使用本地
# # plypath = "D:/RGBD_CAMERA/python_3d_process/PCDPointCloud/fragment.pcd"
# # pcd = o3d.io.read_point_cloud(plypath) # path为文件路径
# plane_model, inliers = pcd.segment_plane(distance_threshold=0.01,
# ransac_n=3,
# num_iterations=1000)
# [a, b, c, d] = plane_model
# print(f"Plane equation: {a:.2f}x + {b:.2f}y + {c:.2f}z + {d:.2f} = 0")
# inlier_cloud = pcd.select_by_index(inliers)
# inlier_cloud.paint_uniform_color([1.0, 0, 0])
# outlier_cloud = pcd.select_by_index(inliers, invert=True)
# o3d.visualization.draw_geometries([inlier_cloud, outlier_cloud],
# zoom=0.8,
# front=[-0.4999, -0.1659, -0.8499],
# lookat=[2.1813, 2.0619, 2.0999],
# up=[0.1204, -0.9852, 0.1215])
#其它例子
plypath = "D:/RGBD_CAMERA/python_3d_process/1_hezi.pcd"
pcd = o3d.io.read_point_cloud(plypath) # path为文件路径
pcd = pcd.remove_non_finite_points(True, False)#剔除无效值
pcd.paint_uniform_color([0.5, 0.5, 0.5])
plane_model, inliers = pcd.segment_plane(distance_threshold=1,
ransac_n=3,
num_iterations=1000)
[a, b, c, d] = plane_model
print(f"Plane equation: {a:.2f}x + {b:.2f}y + {c:.2f}z + {d:.2f} = 0")
inlier_cloud = pcd.select_by_index(inliers)
inlier_cloud.paint_uniform_color([1.0, 0, 0])
outlier_cloud = pcd.select_by_index(inliers, invert=True)
o3d.visualization.draw_geometries([inlier_cloud, outlier_cloud],
zoom=0.8,
front=[-0.4999, -0.1659, -0.8499],
lookat=[2.1813, 2.0619, 2.0999],
up=[0.1204, -0.9852, 0.1215]) （6）测试结果
2.4 隐点去除（Hidden point removal）
（1）函数
hidden_point_removal --- 隐点去除
另外这个例子还用到了
sample_points_poisson_disk -- 泊松磁盘采样
（2）功能说明
想象一下，您想从给定的视点渲染一个点云，但是来自背景的点由于没有被其他点遮挡而泄漏到前景中。为此，我们可以应用隐点去除算法。在Open3D中，[Katz2007]实现了从给定视图逼近点云可见性的方法，而无需进行曲面重构或正常估计
也就是，在特定视角看不到的就不渲染了
（3）本次测试用到的漂亮的模型如下
import open3d as o3d
import numpy as np
from matplotlib import pyplot as plt
# # 初次使用下载模型
print("Convert mesh to a point cloud and estimate dimensions")
armadillo = o3d.data.ArmadilloMesh()
pcd = o3d.io.read_point_cloud(armadillo.path) # path为文件路径
# # 或者直接使用本地
# # plypath = "D:/RGBD_CAMERA/python_3d_process/ArmadilloMesh/ArmadilloMesh.ply"
# # pcd = o3d.io.read_point_cloud(plypath) # path为文件路径
o3d.visualization.draw_geometries([pcd])
（4）测试代码
注意：求了点云值的二范数
np.linalg.norm(求范数)
参考说明地址
np.linalg.norm(求范数)_清晨~的博客-CSDN博客_np.linalg.norm
import open3d as o3d
import numpy as np
from matplotlib import pyplot as plt
# # 初次使用下载模型
print("Convert mesh to a point cloud and estimate dimensions")
armadillo = o3d.data.ArmadilloMesh()
# # pcd = o3d.io.read_point_cloud(armadillo.path) # path为文件路径
# # 或者直接使用本地
# # plypath = "D:/RGBD_CAMERA/python_3d_process/ArmadilloMesh/ArmadilloMesh.ply"
# # pcd = o3d.io.read_point_cloud(plypath) # path为文件路径
# # o3d.visualization.draw_geometries([pcd])
mesh = o3d.io.read_triangle_mesh(armadillo.path)
mesh.compute_vertex_normals()
pcd = mesh.sample_points_poisson_disk(5000)
diameter = np.linalg.norm(
np.asarray(pcd.get_max_bound()) - np.asarray(pcd.get_min_bound()))
o3d.visualization.draw_geometries([pcd])
注意：get_max_bound()和get_min_bound()分别取得体素边界的最大坐标和最小坐标 ，待有空确认？？？
（4）测试结果
（5）设定摄像头视角，去除隐点
import open3d as o3d
import numpy as np
from matplotlib import pyplot as plt
# # 初次使用下载模型
print("Convert mesh to a point cloud and estimate dimensions")
armadillo = o3d.data.ArmadilloMesh()
# # pcd = o3d.io.read_point_cloud(armadillo.path) # path为文件路径
# # 或者直接使用本地
# # plypath = "D:/RGBD_CAMERA/python_3d_process/ArmadilloMesh/ArmadilloMesh.ply"
# # pcd = o3d.io.read_point_cloud(plypath) # path为文件路径
# # o3d.visualization.draw_geometries([pcd])
mesh = o3d.io.read_triangle_mesh(armadillo.path)
mesh.compute_vertex_normals()
pcd = mesh.sample_points_poisson_disk(5000)
diameter = np.linalg.norm(
np.asarray(pcd.get_max_bound()) - np.asarray(pcd.get_min_bound()))
# o3d.visualization.draw_geometries([pcd])
print("Define parameters used for hidden_point_removal")
camera = [0, 0, diameter]
radius = diameter * 100
print("Get all points that are visible from given view point")
_, pt_map = pcd.hidden_point_removal(camera, radius)
print("Visualize result")
pcd = pcd.select_by_index(pt_map)
o3d.visualization.draw_geometries([pcd])
2.5 下采样的其它几个方法
几个函数
voxel_down_sample(voxel_size=0.05) --上一个教程见了，体素下采样，设置体素大小后，将包含在每个体素内的所有点平均计算得到一个点，从而实现下采样 sample_points_poisson_disk -- 泊松磁盘采样，下采样的一个方法uniform_down_sample(every_k_points=5) -- 通过收集每n个点来对点云进行下采样
2.6 点云离群值去除（Point cloud outlier removal）
注意：和我的第2个教程内容基本差不多
（2）点云库处理学习——剔除点云值_chencaw的博客-CSDN博客
官网教程的位置
Point cloud — Open3D 0.16.0 documentation
（1）准备相应数据
import open3d as o3d
import numpy as np
from matplotlib import pyplot as plt
print("Load a ply point cloud, print it, and render it")
sample_pcd_data = o3d.data.PCDPointCloud()
pcd = o3d.io.read_point_cloud(sample_pcd_data.path)
o3d.visualization.draw_geometries([pcd],
zoom=0.3412,
front=[0.4257, -0.2125, -0.8795],
lookat=[2.6172, 2.0475, 1.532],
up=[-0.0694, -0.9768, 0.2024])
print("Downsample the point cloud with a voxel of 0.02")
voxel_down_pcd = pcd.voxel_down_sample(voxel_size=0.02)
o3d.visualization.draw_geometries([voxel_down_pcd],
zoom=0.3412,
front=[0.4257, -0.2125, -0.8795],
lookat=[2.6172, 2.0475, 1.532],
up=[-0.0694, -0.9768, 0.2024]) 
（2）通过uniform_down_sample方法下采样
2.6.1统计剔除函数（Statistical outlier removal）
（1）说明
Statistical_outlier_removal删除距离其邻居更远的点。它接受两个输入参数:
- Nb_neighbors，它指定计算给定点的平均距离时要考虑多少个邻居。
- Std_ratio，它允许基于跨点云平均距离的标准差设置阈值水平。这个数字越低，过滤器的强度就越大。
（2）测试例子
import open3d as o3d
import numpy as np
from matplotlib import pyplot as plt
def display_inlier_outlier(cloud, ind):
inlier_cloud = cloud.select_by_index(ind)
outlier_cloud = cloud.select_by_index(ind, invert=True)
print("Showing outliers (red) and inliers (gray): ")
outlier_cloud.paint_uniform_color([1, 0, 0])
inlier_cloud.paint_uniform_color([0.8, 0.8, 0.8])
o3d.visualization.draw_geometries([inlier_cloud, outlier_cloud],
zoom=0.3412,
front=[0.4257, -0.2125, -0.8795],
lookat=[2.6172, 2.0475, 1.532],
up=[-0.0694, -0.9768, 0.2024])
print("Load a ply point cloud, print it, and render it")
sample_pcd_data = o3d.data.PCDPointCloud()
pcd = o3d.io.read_point_cloud(sample_pcd_data.path)
# o3d.visualization.draw_geometries([pcd],
# zoom=0.3412,
# front=[0.4257, -0.2125, -0.8795],
# lookat=[2.6172, 2.0475, 1.532],
# up=[-0.0694, -0.9768, 0.2024])
print("Downsample the point cloud with a voxel of 0.02")
voxel_down_pcd = pcd.voxel_down_sample(voxel_size=0.02)
# o3d.visualization.draw_geometries([voxel_down_pcd],
# zoom=0.3412,
# front=[0.4257, -0.2125, -0.8795],
# lookat=[2.6172, 2.0475, 1.532],
# up=[-0.0694, -0.9768, 0.2024])
print("Every 5th points are selected")
uni_down_pcd = pcd.uniform_down_sample(every_k_points=5)
# o3d.visualization.draw_geometries([uni_down_pcd],
# zoom=0.3412,
# front=[0.4257, -0.2125, -0.8795],
# lookat=[2.6172, 2.0475, 1.532],
# up=[-0.0694, -0.9768, 0.2024])
print("Statistical oulier removal")
cl, ind = voxel_down_pcd.remove_statistical_outlier(nb_neighbors=20,
std_ratio=2.0)
display_inlier_outlier(voxel_down_pcd, ind)
2.6.2半径离群值去除（Radius outlier removal）
（1）说明
Radius_outlier_removal删除在给定球面上几乎没有邻居的点。有两个参数可以用来调整过滤器到你的数据:
- Nb_points，它允许您选择球体应该包含的最小数量的点。
- 半径，它定义了球体的半径，该半径将用于计数相邻的球体。
（2）测试例子
import open3d as o3d
import numpy as np
from matplotlib import pyplot as plt
def display_inlier_outlier(cloud, ind):
inlier_cloud = cloud.select_by_index(ind)
outlier_cloud = cloud.select_by_index(ind, invert=True)
print("Showing outliers (red) and inliers (gray): ")
outlier_cloud.paint_uniform_color([1, 0, 0])
inlier_cloud.paint_uniform_color([0.8, 0.8, 0.8])
o3d.visualization.draw_geometries([inlier_cloud, outlier_cloud],
zoom=0.3412,
front=[0.4257, -0.2125, -0.8795],
lookat=[2.6172, 2.0475, 1.532],
up=[-0.0694, -0.9768, 0.2024])
print("Load a ply point cloud, print it, and render it")
sample_pcd_data = o3d.data.PCDPointCloud()
pcd = o3d.io.read_point_cloud(sample_pcd_data.path)
# o3d.visualization.draw_geometries([pcd],
# zoom=0.3412,
# front=[0.4257, -0.2125, -0.8795],
# lookat=[2.6172, 2.0475, 1.532],
# up=[-0.0694, -0.9768, 0.2024])
print("Downsample the point cloud with a voxel of 0.02")
voxel_down_pcd = pcd.voxel_down_sample(voxel_size=0.02)
# o3d.visualization.draw_geometries([voxel_down_pcd],
# zoom=0.3412,
# front=[0.4257, -0.2125, -0.8795],
# lookat=[2.6172, 2.0475, 1.532],
# up=[-0.0694, -0.9768, 0.2024])
print("Every 5th points are selected")
uni_down_pcd = pcd.uniform_down_sample(every_k_points=5)
# o3d.visualization.draw_geometries([uni_down_pcd],
# zoom=0.3412,
# front=[0.4257, -0.2125, -0.8795],
# lookat=[2.6172, 2.0475, 1.532],
# up=[-0.0694, -0.9768, 0.2024])
print("Statistical oulier removal")
cl, ind = voxel_down_pcd.remove_statistical_outlier(nb_neighbors=20,
std_ratio=2.0)
display_inlier_outlier(voxel_down_pcd, ind)
print("Radius oulier removal")
cl, ind = voxel_down_pcd.remove_radius_outlier(nb_points=16, radius=0.05)
display_inlier_outlier(voxel_down_pcd, ind)
308
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
