core.pose_estimulation.road_split

Classes

RoadSplit 

RoadSplit()

Attributes

cmd1 instance-attribute 
cmd1 = 'cd ./third/CENet'
cmd2 instance-attribute 
cmd2 = f' python infer.py -d ./data -l ./result -m ./model/512-594 -s valid/test'
dis_th instance-attribute 
dis_th = 5
distance_threshold instance-attribute 
distance_threshold = config.common_config.distance_threshold
img_height instance-attribute 
img_height = config.camera_config.img_height
img_width instance-attribute 
img_width = config.camera_config.img_width
num_iterations instance-attribute 
num_iterations = config.common_config.num_iterations
number_of_decimal instance-attribute 
number_of_decimal = config.common_config.number_of_decimal
ransac_n instance-attribute 
ransac_n = config.common_config.ransac_n
road_range instance-attribute 
road_range = config.common_config.road_range
road_split_label_dir instance-attribute 
road_split_label_dir = config.common_config.road_split_label_dir
road_split_pc_dir instance-attribute 
road_split_pc_dir = config.common_config.road_split_pc_dir

Functions

get_pc_road_in_img 
get_pc_road_in_img(pts_img: numpy.ndarray, pts_rect_depth: numpy.ndarray, points: numpy.ndarray) -> numpy.ndarray
Parameters:
  • pts_img (ndarray) –

    图片中道路的点

  • pts_rect_depth (ndarray) –

    深度图中的点

  • points (ndarray) –

    属于路面的点云
Returns:
  • ndarray –

    可以映射到图片的道路点云
split_pcd_road_by_CENet 
split_pcd_road_by_CENet(bg_index: int, bg_pc_path: str, save_road_label_dir: str, log_dir: str) -> Tuple[numpy.ndarray, numpy.ndarray]
Parameters:
  • bg_index (int) –

    背景索引

  • bg_pc_path (str) –

    背景点云存储路径

  • save_road_label_dir (str) –

    保存路面信息标签的目录

  • log_dir (str) –

    保存日志的目录 # :return: 属于路面的点云, 路面标签, 属于非路面的点云, 属于路面的点云的索引
Returns:
  • Tuple[ndarray, ndarray] –

    属于路面的点云, 属于非路面的点云
split_pcd_road_by_RANSAC 
split_pcd_road_by_RANSAC(bg_index: int, bg_pc_path: str, save_road_label_dir: str, log_dir: str) -> Tuple[numpy.ndarray, numpy.ndarray]
Parameters:
  • bg_index (int) –

    背景索引

  • bg_pc_path (str) –

    背景点云存储路径

  • save_road_label_dir (str) –

    保存路面信息标签的目录

  • log_dir (str) –

    保存日志的目录 # :return: 属于路面的点云, 路面标签, 属于非路面的点云, 属于路面的点云的索引
Returns:
  • Tuple[ndarray, ndarray] –

    属于路面的点云, 属于非路面的点云
split_pcd_road_label 
split_pcd_road_label(labels: list) -> Tuple[List, List]

使用点云路面标签获取路面与非路面的袋牛奶索引

Parameters:
  • labels (list) –

    路面标签
Returns:
  • Tuple[List, List] –

    属于路面的点云的索引,属于非路面的点云的索引
trunc_road_pc 
trunc_road_pc(pc_road: numpy.ndarray) -> numpy.ndarray

保留某个范围的路面点云

Parameters:
  • pc_road (ndarray) –

    路面点云
Returns:
  • ndarray –

    范围截取后的路面点云