bag转换成bin文件

 1、首先安装几个必要的库r，创立py文件，bag2bin.py将以下代码填入，python bag2bin.py即可得到bin文件。

import os
import rosbag
import sensor_msgs.point_cloud2 as pc2
import numpy as np

def bag_to_bin(bag_file, output_dir, topic):
    # 创建输出目录
    if not os.path.exists(output_dir):
        os.makedirs(output_dir)
    
    print(f"Reading bag file: {bag_file}")
    with rosbag.Bag(bag_file, 'r') as bag:
        idx = 0
        for topic, msg, t in bag.read_messages(topics=[topic]):
            points = []
            print(f"Reading message {idx} from topic: {topic} at time: {t}")
            for point in pc2.read_points(msg, skip_nans=True):
                points.append([point[0], point[1], point[2], point[3]])
            
            points = np.array(points, dtype=np.float32)
            print(f"Number of points: {points.shape[0]}")

            # 生成输出bin文件路径
            bin_file = os.path.join(output_dir, f'frame_{idx:06d}.bin')
            print(f"Saving to bin file: {bin_file}")
            points.tofile(bin_file)

            idx += 1
    
    print("Done")

# 单个bag文件路径
bag_file = '/media/happy406/046dcd53-ee66-4d3b-b556-5f9559acab84/0513/person_car.bag'
output_folder = '/media/happy406/046dcd53-ee66-4d3b-b556-5f9559acab84/0513/bin/output_frames/'
topic = '/rslidar_points'

# 将单个bag文件转换为多个bin文件
bag_to_bin(bag_file, output_folder, topic)

2、验证得到的bin文件是否有效，使用open3d来验证（pip install open3d），也是创建一个、check.py文件，将以下代码填入

import numpy as np
import open3d as o3d

def check_bin_file(bin_file):
    # 读取bin文件
    points = np.fromfile(bin_file, dtype=np.float32)
    # 检查文件大小是否正确
    if points.size % 4 != 0:
        print("Error: File size is not correct. It should be a multiple of 4 floats.")
        return

    # 重新调整数组形状为(N, 4)
    points = points.reshape((-1, 4))
    # 打印前几个点检查格式
    print("First few points:")
    print(points[:5])
    # 打印点的总数
    print(f"Total points: {points.shape[0]}")

def visualize_bin_file(bin_file):
    # 读取bin文件
    points = np.fromfile(bin_file, dtype=np.float32).reshape((-1, 4))
    print(f"Points loaded: {points.shape[0]}")
    
    # 打印一些点的信息以供调试
    print("First few points:", points[:5])

    # 创建Open3D点云对象
    pcd = o3d.geometry.PointCloud()
    pcd.points = o3d.utility.Vector3dVector(points[:, :3])

    # 打印点云信息供调试
    print(f"Point cloud contains {len(pcd.points)} points")

    # 可视化点云
    o3d.visualization.draw_geometries([pcd])

# 单个bin文件路径
bin_file = '/media/happy406/046dcd53-ee66-4d3b-b556-5f9559acab84/0513/bin/output_frames/frame_000001.bin'

# 检查bin文件内容
check_bin_file(bin_file)

# 可视化点云数据
visualize_bin_file(bin_file)

结果会通过open3d打开刚才的bin文件，进行可视化，如果显示点云，则转换成功