【Python】Ouster激光雷达pcap数据转存为txt

原创已于 2024-04-21 19:36:12 修改 · 667 阅读

4 ·

CC 4.0 BY-SA版权

文章标签：

#python #Ouster激光雷达

于 2024-04-21 18:23:03 首次发布

本文详细描述了如何在没有官方完整数据解析程序的情况下，安装OusterSDK，处理PCAP文件中的Lidar数据，将其转换为CSV格式，包括预计算、数据提取和文件输出的过程。

官方的sdk没有明确的给出清晰完整的数据解析程序，只好自己写一个，如果对您有用希望给一个免费的赞，感谢~

安装SDK：

参考 https://static.ouster.dev/sdk-docs/installation.html

转存TXT程序


from ouster import client
from ouster import pcap
from itertools import islice
import numpy as np
dual = False
start_index = 0
num_scans = 100
metadata_path = '20240420_1336_OS-2-64-U02_992032000485.json'
pcap_path = '20240420_1336_OS-2-64-U02_992032000485.pcap'
with open(metadata_path, 'r') as f:
    metadata = client.SensorInfo(f.read())
    print(metadata)

source = pcap.Pcap(pcap_path, metadata)

# precompute xyzlut to save computation in a loop
xyzlut = client.XYZLut(metadata)

# create an iterator of LidarScans from pcap and bound it if num is specified
scans = iter(client.Scans(source))
# if num_scans is None
scans = islice(scans, start_index, start_index + num_scans)

row_layer = np.fromfunction(lambda i, j: i,
        (metadata.format.pixels_per_column,
            metadata.format.columns_per_frame), dtype=int)
column_layer = np.fromfunction(lambda i, j: j,
        (metadata.format.pixels_per_column,
            metadata.format.columns_per_frame), dtype=int)
column_layer_staggered = client.destagger(metadata, column_layer,
        inverse=True)

idx = None
for idx, scan in enumerate(scans):

    # initialize the field names for csv header
    # if not field_names or not field_fmts:
    #     field_names, field_fmts = get_fields_info(scan)

    # copy per-column timestamps and measurement_ids for each beam
    timestamps = np.tile(scan.timestamp, (scan.h, 1))
    measurement_ids = np.tile(scan.measurement_id, (scan.h, 1))

    # grab channel data
    fields_values = [scan.field(ch) for ch in scan.fields]

    frame = np.dstack((timestamps, row_layer, column_layer_staggered,
        measurement_ids, *fields_values))

    # output points in "image" vs. staggered order
    frame = client.destagger(metadata, frame)

    # destagger XYZ separately since it has a different type
    xyz = xyzlut(scan.field(client.ChanField.RANGE))
    xyz_destaggered = client.destagger(metadata, xyz)

    if dual:
        xyz2 = xyzlut(scan.field(client.ChanField.RANGE2))
        xyz2_destaggered = client.destagger(metadata, xyz2)

        # get all data as one H x W x num fields int64 array for savetxt()
        frame = np.dstack(tuple(map(lambda x: x.astype(object),
            (frame, xyz_destaggered, xyz2_destaggered))))

    else:
        # get all data as one H x W x num fields int64 array for savetxt()
        frame = np.dstack(tuple(map(lambda x: x.astype(object),
            (frame, xyz_destaggered))))

    frame_colmajor = np.swapaxes(frame, 0, 1)

    # write csv out to file
    csv_path = 'Data1/' + str(idx)+'.txt'
    print(f'write frame index #{idx + start_index}, to file: {csv_path}')
    np.savetxt(csv_path,frame_colmajor.reshape(-1, frame.shape[2]) , fmt='%f', delimiter=' ')
    # header = '\n'.join([f'frame num: {idx}', field_names])

    # np.savetxt(csv_path,
    #            frame_colmajor.reshape(-1, frame.shape[2]),
    #            fmt=field_fmts,
    #            delimiter=',',
    #            header=header)