Open3D可视化连续点云帧（From KITTI tracking dataset）

最新推荐文章于 2024-09-10 09:06:39 发布

炒饭小哪吒

最新推荐文章于 2024-09-10 09:06:39 发布

阅读量1.3k

点赞数

文章标签： numpy python 机器学习

本文链接：https://blog.csdn.net/weixin_45738220/article/details/129367033

版权

设置观看视角
逐点上色
可视化连续点云帧

#!/usr/bin/python3
# -*- coding: utf-8 -*- 

import os
import open3d as o3d
import numpy as np
import time


 
def save_view_point(pcd_numpy, filename):
    vis = o3d.visualization.Visualizer()
    vis.create_window()
    pcd = o3d.open3d.geometry.PointCloud()
    pcd.points= o3d.open3d.utility.Vector3dVector(pcd_numpy)
    vis.add_geometry(pcd)
    axis = o3d.geometry.TriangleMesh.create_coordinate_frame(size=1, origin=[0, 0, 0])
    vis.add_geometry(axis)
    vis.run()  # user changes the view and press "q" to terminate
    param = vis.get_view_control().convert_to_pinhole_camera_parameters()
    o3d.io.write_pinhole_camera_parameters(filename, param)
    vis.destroy_window()

def draw_color(color_h,color_l,pcd):
    color_h = np.array(color_h,np.float32).reshape(1,3)
    color_l = np.array(color_l,np.float32).reshape(1,3)
    
    raw_points = np.array(pcd.points).copy()
    hight = raw_points[:,2:]
    hight = np.clip(hight, -3, 1)
    colors = color_l + (hight - (-3)) * (color_h -  color_l)/4.0
    pcd.colors = o3d.utility.Vector3dVector(colors)
    return pcd

def vis_cons(files_dir,vis_detect_result=False):
    files = os.listdir(files_dir)
    pcds = []
    for f in files:
        pcd_path = os.path.join(files_dir,f)
        pcd=o3d.open3d.geometry.PointCloud()#创建点云对象
        raw_point = np.fromfile(pcd_path, dtype=np.float32).reshape(-1, 4)[:,:3]
        pcd.points= o3d.open3d.utility.Vector3dVector(raw_point)#将点云数据转换为Open3d可以直接使用的数据类型
        pcd = draw_color([1.0,0.36,0.2],[1.0,0.96,0.2],pcd)
        pcds.append(pcd)
    #if vis_detect_result:
    #    batch_results = np.load('batch_results.npy',allow_pickle=True)
    
    vis = o3d.visualization.Visualizer()
    vis.create_window()
    opt = vis.get_render_option()
    opt.background_color = np.asarray([0, 0, 0])
    opt.point_size = 1
    opt.show_coordinate_frame = False
    if os.path.exists("viewpoint.json"):
        ctr = vis.get_view_control()
        param = o3d.io.read_pinhole_camera_parameters("viewpoint.json")
        ctr.convert_from_pinhole_camera_parameters(param)

    for i in range(len(pcds)):
        vis.clear_geometries()
        axis = o3d.geometry.TriangleMesh.create_coordinate_frame(size=1, origin=[0, 0, 0])
        vis.add_geometry(axis)
        vis.add_geometry(pcds[i])
        ctr.convert_from_pinhole_camera_parameters(param)
        time.sleep(0.1)
        vis.run()
    vis.destroy_window()


if __name__ == '__main__':
    exp_pcd_file=r"F:\Datasets\PointCloud\KITTI_track\KITTI_tracking\training\velodyne\0001"
    view_pcd = np.fromfile(os.path.join(exp_pcd_file,'000000.bin'), dtype=np.float32).reshape(-1, 4)[:,:3]
    save_view_point(view_pcd, "viewpoint.json")
    vis_cons(exp_pcd_file)