使用open3d分离背景和物体点云（二）

一、代码

Python

import cv2
import open3d as o3d
import matplotlib.pyplot as plt
import numpy as np


def thPlaneSeg(pointcloud):
    pcd_np = np.asarray(pointcloud.points)

    # 设置深度阈值 (假设Z轴是深度轴)
    depth_threshold = 0.196  # 1.0米

    # 应用深度阈值，移除远于阈值的点
    pcd_np_filtered = pcd_np[pcd_np[:, 2] < depth_threshold]

    # 创建一个新的点云
    pcd_filtered = o3d.geometry.PointCloud()
    pcd_filtered.points = o3d.utility.Vector3dVector(pcd_np_filtered)

    # 如果点云有颜色，则也要相应地过滤颜色
    if pointcloud.colors:
        pcd_colors_np = np.asarray(pointcloud.colors)
        pcd_filtered.colors = o3d.utility.Vector3dVector(pcd_colors_np[pcd_np[:, 2] < depth_threshold])

    # 如果点云有法线，则也要相应地过滤法线
    if pointcloud.normals:
        pcd_normals_np = np.asarray(pointcloud.normals)
        pcd_filtered.normals = o3d.utility.Vector3dVector(pcd_normals_np[pcd_np[:, 2] < depth_threshold])
    return