Statistics of radar data in nuscenes data set

• Focus on v1.0-mini Version of radar data reading and statistics ：

from nuscenes.utils.data_classes import RadarPointCloud
import numpy as np
import matplotlib.pyplot as plt


points_list = []
radar_sensor = [ 'RADAR_BACK_LEFT',
  'RADAR_BACK_RIGHT',
  'RADAR_FRONT',
  'RADAR_FRONT_LEFT',
  'RADAR_FRONT_RIGHT']
# RadarPointCloud The interface of is nuscenes Provide 
for sample in nusc.sample:
  for sensor in radar_sensor:
    sensor_token = sample['data'][sensor]
    points = RadarPointCloud.from_file(str(nusc.get_sample_data_path(sensor_token))).points.transpose(1,0).astype(np.float32)
    points_list.append(points)
print('total points {}'.format(len(points_list)))
point_num = 0
min_points = 2000
max_points = -2
for idx, points in enumerate(points_list):
  point_num += points.shape[0]
  min_points = min(points.shape[0], min_points)
  max_points = max(points.shape[0], max_points)
point_num = point_num*1.0 / len(points_list)
print(f'max_points{
      max_points} min_points{
      min_points} average_points{
      point_num}')

• Output

total points 2020
max_points84 min_points0 average_points39.36584158415842

• statistics

data_dis = [num.shape[0] for num in points_list]

plt.figure(figsize=(10,5))
plt.bar(range(len(data_dis)), data_dis)

• summary ：

1. As shown in the above figure, frame by frame , Sensor by sensor statistics , Each sensor is distributed ：radar_front High data density , Above average
2. radar_back The data density of the post radar is high , Even a frame appears 0 Reflection points
3. The overall distribution is uneven , And there are multiple reflection points of 0 The phenomenon of