深入nuScenes数据集(1/6)
in Tutorial with 0 comment
自动驾驶 数据集 nuScenes Python
深入nuScenes数据集(1/6)
in Tutorial with 0 comment

自动驾驶 数据集 nuScenes Python

这是一篇跟着官方教程走的深入文章，官方教程有6篇。现在这篇是 nuscenes_tutorial.ipynb，也就是第一篇，链接：here

6篇教程分别如下：

背景

nuScenes数据集

传感器在采集车上的布置如下图所示：
2023-05-25T13:40:17.png

准备工作

下载数据集

需要先注册，邮箱注册即可，然后进入下载页：https://www.nuscenes.org/download，找到 Full dataset (v1.0)，md5: 791dd9ced556cfa1b425682f177b5d9b，点击 Asia，即可下载，因为是官方是使用了 aws s3 进行存储，如果遇到下载不畅，请自备魔法。

这个 mini dataset，提供了10个场景供大家去探索数据，足够了，不需要下载 whole dataset。

安装devkit

我这边使用 Anaconda，直接启动 jupyter，所以不需要走安装流程

当然也可以查看官方devkit安装教程，链接：here

查看数据集

将下载好的数据集解压到~/data_sets，完整的目录是/Users/lau/data_sets/v1.0-mini，这个目录后面会用到。

压缩包的内容如下：

 ll -a
total 24
drwx------@  8 lau  staff   256B  5 25 22:14 .
drwxr-xr-x   4 lau  staff   128B  5 25 22:25 ..
-rw-rw-r--@  1 lau  staff   133B  3 23  2019 .v1.0-mini.txt
drwxr-xr-x@  6 lau  staff   192B  3 23  2019 maps
drwxrwxr-x@ 14 lau  staff   448B  3 18  2019 samples
drwxrwxr-x@ 14 lau  staff   448B  3 18  2019 sweeps
drwxr-xr-x@ 15 lau  staff   480B  3 23  2019 v1.0-mini

打开map文件夹，可以看到4个地图的图片。
2023-05-25T14:31:11.png

打开samples文件夹，出现了下面的目录结构，就是上面提到的传感器（6个相机、1个激光雷达、5个毫米波雷达）所采集到的信息

ls -1 samples

CAM_BACK
CAM_BACK_LEFT
CAM_BACK_RIGHT
CAM_FRONT
CAM_FRONT_LEFT
CAM_FRONT_RIGHT
LIDAR_TOP
RADAR_BACK_LEFT
RADAR_BACK_RIGHT
RADAR_FRONT
RADAR_FRONT_LEFT
RADAR_FRONT_RIGHT

打开sweeps文件夹，我们会发现其结构和samples文件夹是完全一样的。samples文件夹中存储的信息是较为关键、重要的，而sweeps文件夹中的信息则相对次要。

打开v1.0-mini文件夹，里面是一堆json格式的文件。这些文件描述了nuScenes中使用的数据库模式。所有的注释和元数据（包括校准、地图、车辆坐标等）都存储在一个关系型数据库中。下面列出了数据库表格。每一行都可以通过其唯一的主键标记进行识别。外键（例如sample_token）可用于连接到sample表的标记。

ls -1 v1.0-mini

attribute.json
calibrated_sensor.json
category.json
ego_pose.json
instance.json
log.json
map.json
sample.json
sample_annotation.json
sample_data.json
scene.json
sensor.json
visibility.json

2023-05-25T14:38:33.png

详细可以看：https://github.com/nutonomy/nuscenes-devkit/blob/master/docs/schema_nuscenes.md

阅读数据集

准备

启动 jupyter，创建nuscenes-devkit.ipynb

安装库

 pip install nuscenes-devkit

导入相关模块和数据集

%matplotlib inline
from nuscenes.nuscenes import NuScenes
nusc = NuScenes(version='v1.0-mini', dataroot='/Users/lau/data_sets/v1.0-mini', verbose=True)

执行后，出现下面信息：

======
Loading NuScenes tables for version v1.0-mini...
23 category,
8 attribute,
4 visibility,
911 instance,
12 sensor,
120 calibrated_sensor,
31206 ego_pose,
8 log,
10 scene,
404 sample,
31206 sample_data,
18538 sample_annotation,
4 map,
Done loading in 0.794 seconds.
======
Reverse indexing ...
Done reverse indexing in 0.1 seconds.
======

中文解释如下：
1.日志（log）- 提取数据的日志信息。
2.场景（scene）- 汽车行驶的20秒片段。
3.样本（sample）- 特定时间戳下场景的注释快照。
4.样本数据（sample_data）- 从特定传感器收集的数据。
5.车辆姿态（ego_pose）- 特定时间戳下车辆的姿态。
6.传感器（sensor）- 特定传感器类型。
7.校准传感器（calibrated sensor）- 定义在特定车辆上校准的特定传感器。
8.实例（instance）- 我们观察到的所有对象实例的枚举。
9.类别（category）- 对象类别的分类法（例如车辆、人类）。
11.属性（attribute）- 实例的属性，在类别保持不变的情况下可以改变。
12.可见度（visibility）- 从6个不同摄像机收集的所有图像中可见的像素分数。
13.样本注释（sample_annotation）- 我们感兴趣的对象的注释实例。
14.地图（map）- 以二进制语义掩码的形式存储的地图数据，以俯视图呈现。

场景（scene）

nusc.list_scenes()

# 输出结果
# 每行的意思分别是场景名称、场景中的景物描述、采集时间、场景持续时间、采集地点、场景注释数量
scene-0061, Parked truck, construction, intersectio... [18-07-24 03:28:47]   19s, singapore-onenorth, #anns:4622
scene-0103, Many peds right, wait for turning car, ... [18-08-01 19:26:43]   19s, boston-seaport, #anns:2046
scene-0655, Parking lot, parked cars, jaywalker, be... [18-08-27 15:51:32]   20s, boston-seaport, #anns:2332
scene-0553, Wait at intersection, bicycle, large tr... [18-08-28 20:48:16]   20s, boston-seaport, #anns:1950
scene-0757, Arrive at busy intersection, bus, wait ... [18-08-30 19:25:08]   20s, boston-seaport, #anns:592
scene-0796, Scooter, peds on sidewalk, bus, cars, t... [18-10-02 02:52:24]   20s, singapore-queensto, #anns:708
scene-0916, Parking lot, bicycle rack, parked bicyc... [18-10-08 07:37:13]   20s, singapore-queensto, #anns:2387
scene-1077, Night, big street, bus stop, high speed... [18-11-21 11:39:27]   20s, singapore-hollandv, #anns:890
scene-1094, Night, after rain, many peds, PMD, ped ... [18-11-21 11:47:27]   19s, singapore-hollandv, #anns:1762
scene-1100, Night, peds in sidewalk, peds cross cro... [18-11-21 11:49:47]   19s, singapore-hollandv, #anns:935
my_scene = nusc.scene[0]
my_scene

# 输出结果
# 可以通过 token 获取对应的信息
{'token': 'cc8c0bf57f984915a77078b10eb33198',
 'log_token': '7e25a2c8ea1f41c5b0da1e69ecfa71a2',
 'nbr_samples': 39,
 'first_sample_token': 'ca9a282c9e77460f8360f564131a8af5',
 'last_sample_token': 'ed5fc18c31904f96a8f0dbb99ff069c0',
 'name': 'scene-0061',
 'description': 'Parked truck, construction, intersection, turn left, following a van'}

样本（sample）

先来说说sample和scene的关系，前面说到，每个scene大约持续20s，那sample就是每0.5秒进行一次采样。也可以这样理解sample和scene，sence相当于20s的视频，sample就是每0.5s取一帧的图像。

first_sample_token = my_scene['first_sample_token']  #获取第一个sample的token值
first_sample_token 

# 输出结果
'ca9a282c9e77460f8360f564131a8af5'
my_sample = nusc.get('sample', first_sample_token)
my_sample

# 输出结果
# 结果中包含了传感器采集到的信息、注释信息等等
{'token': 'ca9a282c9e77460f8360f564131a8af5',
 'timestamp': 1532402927647951,
 'prev': '',
 'next': '39586f9d59004284a7114a68825e8eec',
 'scene_token': 'cc8c0bf57f984915a77078b10eb33198',
 'data': {'RADAR_FRONT': '37091c75b9704e0daa829ba56dfa0906',
  'RADAR_FRONT_LEFT': '11946c1461d14016a322916157da3c7d',
  'RADAR_FRONT_RIGHT': '491209956ee3435a9ec173dad3aaf58b',
  'RADAR_BACK_LEFT': '312aa38d0e3e4f01b3124c523e6f9776',
  'RADAR_BACK_RIGHT': '07b30d5eb6104e79be58eadf94382bc1',
  'LIDAR_TOP': '9d9bf11fb0e144c8b446d54a8a00184f',
  'CAM_FRONT': 'e3d495d4ac534d54b321f50006683844',
  'CAM_FRONT_RIGHT': 'aac7867ebf4f446395d29fbd60b63b3b',
  'CAM_BACK_RIGHT': '79dbb4460a6b40f49f9c150cb118247e',
  'CAM_BACK': '03bea5763f0f4722933508d5999c5fd8',
  'CAM_BACK_LEFT': '43893a033f9c46d4a51b5e08a67a1eb7',
  'CAM_FRONT_LEFT': 'fe5422747a7d4268a4b07fc396707b23'},
 'anns': ['ef63a697930c4b20a6b9791f423351da',
  '6b89da9bf1f84fd6a5fbe1c3b236f809',
  ...
  '2bfcc693ae9946daba1d9f2724478fd4']}
nusc.list_sample(my_sample['token'])

# 输出结果
# 列出所有 sample_data 关键帧和样本相关的 sample_annotation
Sample: ca9a282c9e77460f8360f564131a8af5

sample_data_token: 37091c75b9704e0daa829ba56dfa0906, mod: radar, channel: RADAR_FRONT
sample_data_token: 11946c1461d14016a322916157da3c7d, mod: radar, channel: RADAR_FRONT_LEFT
sample_data_token: 491209956ee3435a9ec173dad3aaf58b, mod: radar, channel: RADAR_FRONT_RIGHT
sample_data_token: 312aa38d0e3e4f01b3124c523e6f9776, mod: radar, channel: RADAR_BACK_LEFT
sample_data_token: 07b30d5eb6104e79be58eadf94382bc1, mod: radar, channel: RADAR_BACK_RIGHT
sample_data_token: 9d9bf11fb0e144c8b446d54a8a00184f, mod: lidar, channel: LIDAR_TOP
sample_data_token: e3d495d4ac534d54b321f50006683844, mod: camera, channel: CAM_FRONT
sample_data_token: aac7867ebf4f446395d29fbd60b63b3b, mod: camera, channel: CAM_FRONT_RIGHT
sample_data_token: 79dbb4460a6b40f49f9c150cb118247e, mod: camera, channel: CAM_BACK_RIGHT
sample_data_token: 03bea5763f0f4722933508d5999c5fd8, mod: camera, channel: CAM_BACK
sample_data_token: 43893a033f9c46d4a51b5e08a67a1eb7, mod: camera, channel: CAM_BACK_LEFT
sample_data_token: fe5422747a7d4268a4b07fc396707b23, mod: camera, channel: CAM_FRONT_LEFT

sample_annotation_token: ef63a697930c4b20a6b9791f423351da, category: human.pedestrian.adult
...
sample_annotation_token: 2bfcc693ae9946daba1d9f2724478fd4, category: movable_object.barrier

样本数据（sample_data）

获取样本数据

my_sample['data']

# 输出结果
{'RADAR_FRONT': '37091c75b9704e0daa829ba56dfa0906',
 'RADAR_FRONT_LEFT': '11946c1461d14016a322916157da3c7d',
 'RADAR_FRONT_RIGHT': '491209956ee3435a9ec173dad3aaf58b',
 'RADAR_BACK_LEFT': '312aa38d0e3e4f01b3124c523e6f9776',
 'RADAR_BACK_RIGHT': '07b30d5eb6104e79be58eadf94382bc1',
 'LIDAR_TOP': '9d9bf11fb0e144c8b446d54a8a00184f',
 'CAM_FRONT': 'e3d495d4ac534d54b321f50006683844',
 'CAM_FRONT_RIGHT': 'aac7867ebf4f446395d29fbd60b63b3b',
 'CAM_BACK_RIGHT': '79dbb4460a6b40f49f9c150cb118247e',
 'CAM_BACK': '03bea5763f0f4722933508d5999c5fd8',
 'CAM_BACK_LEFT': '43893a033f9c46d4a51b5e08a67a1eb7',
 'CAM_FRONT_LEFT': 'fe5422747a7d4268a4b07fc396707b23'}

获取 LIDAR_TOP 的元数据

sensor = 'LIDAR_TOP'
lidar_top = nusc.get('sample_data', my_sample['data'][sensor])
lidar_top

# 输出结果
{'token': '9d9bf11fb0e144c8b446d54a8a00184f',
 'sample_token': 'ca9a282c9e77460f8360f564131a8af5',
 'ego_pose_token': '9d9bf11fb0e144c8b446d54a8a00184f',
 'calibrated_sensor_token': 'a183049901c24361a6b0b11b8013137c',
 'timestamp': 1532402927647951,
 'fileformat': 'pcd',
 'is_key_frame': True,
 'height': 0,
 'width': 0,
 'filename': 'samples/LIDAR_TOP/n015-2018-07-24-11-22-45+0800__LIDAR_TOP__1532402927647951.pcd.bin',
 'prev': '',
 'next': '0cedf1d2d652468d92d23491136b5d15',
 'sensor_modality': 'lidar',
 'channel': 'LIDAR_TOP'}

可视化展示数据：

nusc.render_sample_data(lidar_top_data['token'])

2023-05-25T15:59:59.png

这里只展示了 LIDAR_TOP ，即车顶的激光雷达的可视化结果，可视化其他传感器的方法也类似的。

样本注释（sample_annotation）

获取样本注释

my_annotation_token = my_sample['anns'][18]
my_annotation_metadata = nusc.get('sample_annotation', my_annotation_token)
my_annotation_metadata

# 输出结果
{'token': '83d881a6b3d94ef3a3bc3b585cc514f8',
 'sample_token': 'ca9a282c9e77460f8360f564131a8af5',
 'instance_token': 'e91afa15647c4c4994f19aeb302c7179',
 'visibility_token': '4',
 'attribute_tokens': ['58aa28b1c2a54dc88e169808c07331e3'],
 'translation': [409.989, 1164.099, 1.623],
 'size': [2.877, 10.201, 3.595],
 'rotation': [-0.5828819500503033, 0.0, 0.0, 0.812556848660791],
 'prev': '',
 'next': 'f3721bdfd7ee4fd2a4f94874286df471',
 'num_lidar_pts': 495,
 'num_radar_pts': 13,
 'category_name': 'vehicle.truck'}

可视化展示数据：

nusc.render_annotation(my_annotation_token)

2023-05-25T16:04:37.png

实例（instance）

获取第599个实例

my_instance = nusc.instance[599]
my_instance

# 输出结果
{'token': '9cba9cd8af85487fb010652c90d845b5',
 'category_token': 'fedb11688db84088883945752e480c2c',
 'nbr_annotations': 16,
 'first_annotation_token': '77afa772cb4a4e5c8a5a53f2019bdba0',
 'last_annotation_token': '6fed6d902e5e487abb7444f62e1a2341'}

可视化这个实例

instance_token = my_instance['token']
nusc.render_instance(instance_token)

2023-05-25T16:11:56.png

可视化这个实例第一个注释

print("First annotated sample of this instance:")
nusc.render_annotation(my_instance['first_annotation_token'])

2023-05-25T16:12:19.png

可视化这个实例最后一个注释

print("Last annotated sample of this instance")
nusc.render_annotation(my_instance['last_annotation_token'])

2023-05-25T16:12:37.png

这里的注释可以理解为对这实例的追踪

类别（category）

获取类别

nusc.list_categories()

# 输出结果
Category stats for split v1.0-mini:
human.pedestrian.adult      n= 4765, width= 0.68±0.11, len= 0.73±0.17, height= 1.76±0.12, lw_aspect= 1.08±0.23
human.pedestrian.child      n=   46, width= 0.46±0.08, len= 0.45±0.09, height= 1.37±0.06, lw_aspect= 0.97±0.05
human.pedestrian.constructi n=  193, width= 0.69±0.07, len= 0.74±0.12, height= 1.78±0.05, lw_aspect= 1.07±0.16
human.pedestrian.personal_m n=   25, width= 0.83±0.00, len= 1.28±0.00, height= 1.87±0.00, lw_aspect= 1.55±0.00
human.pedestrian.police_off n=   11, width= 0.59±0.00, len= 0.47±0.00, height= 1.81±0.00, lw_aspect= 0.80±0.00
movable_object.barrier      n= 2323, width= 2.32±0.49, len= 0.61±0.11, height= 1.06±0.10, lw_aspect= 0.28±0.09
movable_object.debris       n=   13, width= 0.43±0.00, len= 1.43±0.00, height= 0.46±0.00, lw_aspect= 3.35±0.00
movable_object.pushable_pul n=   82, width= 0.51±0.06, len= 0.79±0.10, height= 1.04±0.20, lw_aspect= 1.55±0.18
movable_object.trafficcone  n= 1378, width= 0.47±0.14, len= 0.45±0.07, height= 0.78±0.13, lw_aspect= 0.99±0.12
static_object.bicycle_rack  n=   54, width= 2.67±1.46, len=10.09±6.19, height= 1.40±0.00, lw_aspect= 5.97±4.02
vehicle.bicycle             n=  243, width= 0.64±0.12, len= 1.82±0.14, height= 1.39±0.34, lw_aspect= 2.94±0.41
vehicle.bus.bendy           n=   57, width= 2.83±0.09, len= 9.23±0.33, height= 3.32±0.07, lw_aspect= 3.27±0.22
vehicle.bus.rigid           n=  353, width= 2.95±0.26, len=11.46±1.79, height= 3.80±0.62, lw_aspect= 3.88±0.57
vehicle.car                 n= 7619, width= 1.92±0.16, len= 4.62±0.36, height= 1.69±0.21, lw_aspect= 2.41±0.18
vehicle.construction        n=  196, width= 2.58±0.35, len= 5.57±1.57, height= 2.38±0.33, lw_aspect= 2.18±0.62
vehicle.motorcycle          n=  471, width= 0.68±0.21, len= 1.95±0.38, height= 1.47±0.20, lw_aspect= 3.00±0.62
vehicle.trailer             n=   60, width= 2.28±0.08, len=10.14±5.69, height= 3.71±0.27, lw_aspect= 4.37±2.41
vehicle.truck               n=  649, width= 2.35±0.34, len= 6.50±1.56, height= 2.62±0.68, lw_aspect= 2.75±0.37

获取第i个类别的信息

nusc.category[9]

# 输出结果
{'token': 'dfd26f200ade4d24b540184e16050022',
 'name': 'vehicle.motorcycle',
 'description': 'Gasoline or electric powered 2-wheeled vehicle designed to move rapidly (at the speed of standard cars) on the road surface. This category includes all motorcycles, vespas and scooters.'}

属性（attribute）

获取属性

nusc.list_attributes()

# 输出结果
cycle.with_rider: 305
cycle.without_rider: 434
pedestrian.moving: 3875
pedestrian.sitting_lying_down: 111
pedestrian.standing: 1029
vehicle.moving: 2715
vehicle.parked: 4674
vehicle.stopped: 1545

在一个场景中，实例的属性是会变的，下面展示行人的属性的变化情况：从移动变为站立

my_instance = nusc.instance[27]
first_token = my_instance['first_annotation_token']
last_token = my_instance['last_annotation_token']
nbr_samples = my_instance['nbr_annotations']
current_token = first_token

i = 0
found_change = False
while current_token != last_token:
    current_ann = nusc.get('sample_annotation', current_token)
    current_attr = nusc.get('attribute', current_ann['attribute_tokens'][0])['name']
    
    if i == 0:
        pass
    elif current_attr != last_attr:
        print("Changed from `{}` to `{}` at timestamp {} out of {} annotated timestamps".format(last_attr, current_attr, i, nbr_samples))
        found_change = True

    next_token = current_ann['next']
    current_token = next_token
    last_attr = current_attr
    i += 1

# 输出结果
Changed from `pedestrian.moving` to `pedestrian.standing` at timestamp 21 out of 39 annotated timestamps

可见度（visibility）

获取可见度分级

nusc.visibility

# 输出结果
[{'description': 'visibility of whole object is between 0 and 40%',
  'token': '1',
  'level': 'v0-40'},
 {'description': 'visibility of whole object is between 40 and 60%',
  'token': '2',
  'level': 'v40-60'},
 {'description': 'visibility of whole object is between 60 and 80%',
  'token': '3',
  'level': 'v60-80'},
 {'description': 'visibility of whole object is between 80 and 100%',
  'token': '4',
  'level': 'v80-100'}]

获取一个 v80-100 的样本注释

anntoken = 'a7d0722bce164f88adf03ada491ea0ba'
visibility_token = nusc.get('sample_annotation', anntoken)['visibility_token']

print("Visibility: {}".format(nusc.get('visibility', visibility_token)))
nusc.render_annotation(anntoken)

Visibility: {'description': 'visibility of whole object is between 80 and 100%', 'token': '4', 'level': 'v80-100'}
2023-05-25T16:25:08.png

传感器（sensor）

获取传感器

nusc.sensor

# 输出结果
[{'token': '725903f5b62f56118f4094b46a4470d8',
  'channel': 'CAM_FRONT',
  'modality': 'camera'},
 {'token': 'ce89d4f3050b5892b33b3d328c5e82a3',
  'channel': 'CAM_BACK',
  'modality': 'camera'},
 {'token': 'a89643a5de885c6486df2232dc954da2',
  'channel': 'CAM_BACK_LEFT',
  'modality': 'camera'},
 {'token': 'ec4b5d41840a509984f7ec36419d4c09',
  'channel': 'CAM_FRONT_LEFT',
  'modality': 'camera'},
 {'token': '2f7ad058f1ac5557bf321c7543758f43',
  'channel': 'CAM_FRONT_RIGHT',
  'modality': 'camera'},
 {'token': 'ca7dba2ec9f95951bbe67246f7f2c3f7',
  'channel': 'CAM_BACK_RIGHT',
  'modality': 'camera'},
 {'token': 'dc8b396651c05aedbb9cdaae573bb567',
  'channel': 'LIDAR_TOP',
  'modality': 'lidar'},
 {'token': '47fcd48f71d75e0da5c8c1704a9bfe0a',
  'channel': 'RADAR_FRONT',
  'modality': 'radar'},
 {'token': '232a6c4dc628532e81de1c57120876e9',
  'channel': 'RADAR_FRONT_RIGHT',
  'modality': 'radar'},
 {'token': '1f69f87a4e175e5ba1d03e2e6d9bcd27',
  'channel': 'RADAR_FRONT_LEFT',
  'modality': 'radar'},
 {'token': 'df2d5b8be7be55cca33c8c92384f2266',
  'channel': 'RADAR_BACK_LEFT',
  'modality': 'radar'},
 {'token': '5c29dee2f70b528a817110173c2e71b9',
  'channel': 'RADAR_BACK_RIGHT',
  'modality': 'radar'}]

因sample_data中就存储着传感器的信息，因此可以通过nusc.sample_data[i]来获取传感器的信息

nusc.sample_data[10]


# 输出结果
{'token': '2ecfec536d984fb491098c9db1404117',
 'sample_token': '356d81f38dd9473ba590f39e266f54e5',
 'ego_pose_token': '2ecfec536d984fb491098c9db1404117',
 'calibrated_sensor_token': 'f4d2a6c281f34a7eb8bb033d82321f79',
 'timestamp': 1532402928269133,
 'fileformat': 'pcd',
 'is_key_frame': False,
 'height': 0,
 'width': 0,
 'filename': 'sweeps/RADAR_FRONT/n015-2018-07-24-11-22-45+0800__RADAR_FRONT__1532402928269133.pcd',
 'prev': 'b933bbcb4ee84a7eae16e567301e1df2',
 'next': '79ef24d1eba84f5abaeaf76655ef1036',
 'sensor_modality': 'radar',
 'channel': 'RADAR_FRONT'}

校准传感器（calibrated sensor）

获取某传感器的校准信息

nusc.calibrated_sensor[0]

# 输出结果
{'token': 'f4d2a6c281f34a7eb8bb033d82321f79',
 'sensor_token': '47fcd48f71d75e0da5c8c1704a9bfe0a',
 'translation': [3.412, 0.0, 0.5],
 'rotation': [0.9999984769132877, 0.0, 0.0, 0.0017453283658983088],
 'camera_intrinsic': []}

车辆姿态（ego_pose）

nusc.ego_pose[0]

# 输出结果
{'token': '5ace90b379af485b9dcb1584b01e7212',
 'timestamp': 1532402927814384,
 'rotation': [0.5731787718287827,
  -0.0015811634307974854,
  0.013859363182046986,
  -0.8193116095230444],
 'translation': [410.77878632230204, 1179.4673290964536, 0.0]}

日志（log）

获取日志的数量

print("Number of `logs` in our loaded database: {}".format(len(nusc.log)))

# 输出结果
Number of `logs` in our loaded database: 8

获取某个日志

nusc.log[0]

# 输出结果
{'token': '7e25a2c8ea1f41c5b0da1e69ecfa71a2',
 'logfile': 'n015-2018-07-24-11-22-45+0800',
 'vehicle': 'n015',
 'date_captured': '2018-07-24',
 'location': 'singapore-onenorth',
 'map_token': '53992ee3023e5494b90c316c183be829'}

地图（map）

获取地图的数量

print("There are {} maps masks in the loaded dataset".format(len(nusc.map)))

# 输出结果
There are 4 maps masks in the loaded dataset

获取某个地图

nusc.map[0]

{'category': 'semantic_prior',
 'token': '53992ee3023e5494b90c316c183be829',
 'filename': 'maps/53992ee3023e5494b90c316c183be829.png',
 'log_tokens': ['0986cb758b1d43fdaa051ab23d45582b',
  '1c9b302455ff44a9a290c372b31aa3ce',
  'e60234ec7c324789ac7c8441a5e49731',
  '46123a03f41e4657adc82ed9ddbe0ba2',
  'a5bb7f9dd1884f1ea0de299caefe7ef4',
  'bc41a49366734ebf978d6a71981537dc',
  'f8699afb7a2247e38549e4d250b4581b',
  'd0450edaed4a46f898403f45fa9e5f0d',
  'f38ef5a1e9c941aabb2155768670b92a',
  '7e25a2c8ea1f41c5b0da1e69ecfa71a2',
  'ddc03471df3e4c9bb9663629a4097743',
  '31e9939f05c1485b88a8f68ad2cf9fa4',
  '783683d957054175bda1b326453a13f4',
  '343d984344e440c7952d1e403b572b2a',
  '92af2609d31445e5a71b2d895376fed6',
  '47620afea3c443f6a761e885273cb531',
  'd31dc715d1c34b99bd5afb0e3aea26ed',
  '34d0574ea8f340179c82162c6ac069bc',
  'd7fd2bb9696d43af901326664e42340b',
  'b5622d4dcb0d4549b813b3ffb96fbdc9',
  'da04ae0b72024818a6219d8dd138ea4b',
  '6b6513e6c8384cec88775cae30b78c0e',
  'eda311bda86f4e54857b0554639d6426',
  'cfe71bf0b5c54aed8f56d4feca9a7f59',
  'ee155e99938a4c2698fed50fc5b5d16a',
  '700b800c787842ba83493d9b2775234a'],
 'mask': <nuscenes.utils.map_mask.MapMask at 0x7fb4c3e8a9b0>}

进阶

播放传感器

播放其中一个传感器

my_scene_token = nusc.field2token('scene', 'name', 'scene-0061')[0]
nusc.render_scene_channel(my_scene_token, 'CAM_FRONT')

2023-05-25T16:53:27.png

播放所有传感器

nusc.render_scene(my_scene_token)

2023-05-25T16:54:36.png

绘制车辆行驶地图

nusc.render_egoposes_on_map(log_location='singapore-onenorth')

# 输出结果
Adding ego poses to map...
100%|█████████████████████████████████████████████| 1/1 [00:00<00:00, 37.87it/s]
Creating plot...

2023-05-25T16:47:34.png

参考

本文由 Chakhsu Lau 创作，采用 知识共享署名4.0 国际许可协议进行许可。
本站文章除注明转载/出处外，均为本站原创或翻译，转载前请务必署名。

← → ↓ ↑
Responses