Open3D点云基础操作报错退出（错误码3221225477）求助

问题现象 #

• 点云可正常可视化，但执行transform、paint_uniform_color等操作时，程序直接以错误码3221225477退出
• 向o3d.visualization.draw_geometries传递相机参数时，该函数无法正常工作
• 注释掉source.transform(transformation)、source.paint_uniform_color及target.paint_uniform_color相关代码后，程序可正常运行并可视化点云

环境信息 #

• 系统：Windows 10 64位
• Python版本：3.11.9
• 处理器：x64架构
• Open3D版本：0.18.0，通过pip安装

已尝试的解决方法 #

• 在虚拟环境中运行代码
• 使用以下命令重新安装Open3D：
  • pip install open3d
  • pip3 install open3d
  • pip install --user open3d
    以上操作均未解决问题

运行代码 #

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

def draw_registration_result(source, target, transformation):
    #source_temp = copy.deep_copy(source)
    #target_temp = copy.deepcopy(target)
    source.paint_uniform_color([1, 0.706, 0])
    target.paint_uniform_color([0, 0.651, 0.929])
    #source.transform(transformation)
    o3d.visualization.draw_geometries([source, target])

def preprocess_point_cloud(pcd, voxel_size):
    print(":: Downsample with a voxel size %.3f." % voxel_size)
    pcd_down = pcd.voxel_down_sample(voxel_size = voxel_size)
    radius_normal = voxel_size * 2
    radius_feature = voxel_size * 5
    print(":: Estimate normal with search radius %.3f." % radius_normal)
    pcd_down.estimate_normals(o3d.geometry.KDTreeSearchParamHybrid(radius = radius_normal, max_nn = 30))
    print(":: Compute FPFH feature with search radius %.3f." % radius_feature)
    pcd_fpfh = o3d.pipelines.registration.compute_fpfh_feature(pcd_down, 
                                                               o3d.geometry.KDTreeSearchParamHybrid(radius = radius_feature, max_nn = 100))
    return pcd_down, pcd_fpfh

def prepare_dataset(voxel_size):
    print(":: Load two point clouds and disturb initial pose.")
    demo_icp_pcds = o3d.data.DemoICPPointClouds()
    source = o3d.io.read_point_cloud(demo_icp_pcds.paths[0])
    target = o3d.io.read_point_cloud(demo_icp_pcds.paths[1])
    transformation = np.asarray([[0.0, 0.0, 1.0, 0.0], [1.0, 0.0, 0.0, 0.0],
                                [0.0, 1.0, 0.0, 0.0], [0.0, 0.0, 0.0, 1.0]])
    source.transform(transformation)
    draw_registration_result(source, target, np.identity(4))
    source_down, source_fpfh = preprocess_point_cloud(source, voxel_size)
    target_down, target_fpfh = preprocess_point_cloud(target, voxel_size)
    return source, target, source_down, target_down, source_fpfh, target_fpfh

voxel_size = 0.05
source, target, source_down, target_down, source_fpfh, target_fpfh = prepare_dataset(voxel_size)

报错信息 #

[Running] python -u "d:\python\Open3D\Practice.py"
:: Load two point clouds and disturb initial pose.

[Done] exited with code=3221225477 in 2.787 seconds

内容的提问来源于stack exchange，提问作者Abhishek Mahato