用Python写激光雷达点云pcd障碍物识别并用open3d可视化

要实现激光雷达点云pcd障碍物识别并用open3d可视化，可以按照以下步骤进行：

1. 安装open3d和numpy库：

pip install open3d numpy

2. 读取pcd文件：

import open3d as o3d

pcd = o3d.io.read_point_cloud("example.pcd")

3. 对点云进行下采样，以加快处理速度：

downpcd = pcd.voxel_down_sample(voxel_size=0.05)

4. 对下采样后的点云进行平滑处理，以去除噪声：

downpcd.estimate_normals()
downpcd.paint_uniform_color([1, 0.706, 0])
o3d.visualization.draw_geometries([downpcd])

5. 对处理后的点云进行聚类，找出障碍物：

with o3d.utility.VerbosityContextManager(
        o3d.utility.VerbosityLevel.Debug) as cm:
    labels = np.array(downpcd.cluster_dbscan(eps=0.5, 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
downpcd.colors = o3d.utility.Vector3dVector(colors[:, :3])
o3d.visualization.draw_geometries([downpcd])

6. 可以将障碍物的点云保存成独立的pcd文件：

for i in range(max_label + 1):
    indices = np.where(labels == i)[0]
    cloud_cluster = downpcd.select_down_sample(indices)
    o3d.io.write_point_cloud(f"cluster{i}.pcd", cloud_cluster)

7. 最后，可以用open3d可视化显示障碍物识别结果：

o3d.visualization.draw_geometries([downpcd])

完整代码如下：

import numpy as np
import open3d as o3d
import matplotlib.pyplot as plt

# 读取pcd文件
pcd = o3d.io.read_point_cloud("example.pcd")

# 对点云进行下采样
downpcd = pcd.voxel_down_sample(voxel_size=0.05)

# 对下采样后的点云进行平滑处理
downpcd.estimate_normals()
downpcd.paint_uniform_color([1, 0.706, 0])
o3d.visualization.draw_geometries([downpcd])

# 对处理后的点云进行聚类，找出障碍物
with o3d.utility.VerbosityContextManager(
        o3d.utility.VerbosityLevel.Debug) as cm:
    labels = np.array(downpcd.cluster_dbscan(eps=0.5, 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
downpcd.colors = o3d.utility.Vector3dVector(colors[:, :3])
o3d.visualization.draw_geometries([downpcd])

# 将障碍物的点云保存成独立的pcd文件
for i in range(max_label + 1):
    indices = np.where(labels == i)[0]
    cloud_cluster = downpcd.select_down_sample(indices)
    o3d.io.write_point_cloud(f"cluster{i}.pcd", cloud_cluster)

# 用open3d可视化显示障碍物识别结果
o3d.visualization.draw_geometries([downpcd])