utils.Utils_box

Classes

UtilsBox

Functions

change_box3d staticmethod 
change_box3d(box3d)
Parameters:
  • box3d –

    插入物体box3d
Returns:
  • –

    矫正后的box3d,box3d的欧拉旋转角
convert_box3d2corner_box staticmethod 
convert_box3d2corner_box(box3d: o3d.geometry.OrientedBoundingBox) -> numpy.ndarray
Parameters:
  • box3d (OrientedBoundingBox) –

    物体box3d
Returns:
  • ndarray –

    物体box3d的八个角
convert_corner_box2box3d staticmethod 
convert_corner_box2box3d(corner: numpy.ndarray) -> o3d.geometry.OrientedBoundingBox
Parameters:
  • corner (ndarray) –

    物体box3d的八个角
Returns:
  • OrientedBoundingBox –

    物体box3d
convert_label_box3d2corner_box staticmethod 
convert_label_box3d2corner_box(label_boxes3d: numpy.ndarray) -> numpy.ndarray
Parameters:
  • label_boxes3d (ndarray) –

    (N, 7) [x, y, z, dx, dy, dz, heading], (x, y, z) 是物体边界框中心
Returns:
  • ndarray –

    物体点云边界框的八个角
get_box2d_center staticmethod 
get_box2d_center(box: List[float]) -> List[int]
Parameters:
  • box (List[float]) –

    坐标范围
Returns:
  • List[int] –

    坐标范围的中心点
get_box2d_from_image staticmethod 
get_box2d_from_image(image_obj: numpy.ndarray, pos_image: List[int]) -> List[float]
Parameters:
  • image_obj (ndarray) –

    物体图片

  • pos_image (List[int]) –

    背景图片中心点坐标
Returns:
  • List[float] –

    物体图片在背景图片上的坐标范围
get_box2d_from_points staticmethod 
get_box2d_from_points(pts: numpy.ndarray) -> List[float]
Parameters:
  • pts (ndarray) –

    物体点云映射到到背景图片的点
Returns:
  • List[float] –

    物体点云映射到背景图片中的坐标范围
get_euler_from_rotate_matrix staticmethod 
get_euler_from_rotate_matrix(R: numpy.ndarray) -> List[float]
Parameters:
  • R (ndarray) –

    旋转矩阵
Returns:
  • List[float] –

    构成旋转矩阵的欧拉角(弧度制)
iou_2d staticmethod 
iou_2d(box0: List[float], box1: List[float]) -> float
Parameters:
  • box0 (List[float]) –

    坐标范围-0

  • box1 (List[float]) –

    坐标范围-1
Returns:
  • float –

    IOU值
rotate_points_along_z staticmethod 
rotate_points_along_z(points: torch.Tensor, angle: torch.Tensor) -> numpy.ndarray
Parameters:
  • points (Tensor) –

    points: (B, N, 3 + C)

  • angle (Tensor) –

    angle: (B), angle along z-axis, angle increases x ==> y
Returns:
  • ndarray –

    旋转后的点云
trunc_box2d_from_img staticmethod 
trunc_box2d_from_img(box: List[float], max_x: int, max_y: int) -> List[float]
Parameters:
  • box (List[float]) –

    物体点云映射到背景图片中的坐标范围

  • max_x (int) –

    背景图片的水平尺寸

  • max_y (int) –

    背景图片的垂直尺寸
Returns:
  • List[float] –

    若坐标范围在背景图片内，无变化；否则，将超出部分截断后返回