Mojo struct
BoundingBox
struct BoundingBox[dtype: DType]
Represents a 2D bounding box for object detection.
The box is stored using two corner points: nw and se.
Note: In this implementation, nw stores the maximum coordinates (max y, max x)
and se stores the minimum coordinates (min y, min x). This differs from the typical
interpretation of "northwest" (usually min x, max y) and "southeast" (usually max x, min y).
This representation allows efficient computation of intersection and union areas.
Fields: nw: Corner storing the maximum coordinates (max y, max x). se: Corner storing the minimum coordinates (min y, min x).
Parameters
- dtype (
DType): The data type for coordinate values.
Fields
- nw (
SIMD[dtype, 2]): - se (
SIMD[dtype, 2]):
Implemented traits
AnyType,
Copyable,
ImplicitlyCopyable,
ImplicitlyDestructible,
Movable
Methods
__init__
__init__(out self, y1: Scalar[dtype], x1: Scalar[dtype], y2: Scalar[dtype], x2: Scalar[dtype])
Initialize a bounding box from two diagonal corner coordinates.
Note: The corners are automatically ordered to ensure nw contains the maximum coordinates and se contains the minimum coordinates.
Args:
- y1 (
Scalar[dtype]): Y-coordinate of first corner. - x1 (
Scalar[dtype]): X-coordinate of first corner. - y2 (
Scalar[dtype]): Y-coordinate of second corner. - x2 (
Scalar[dtype]): X-coordinate of second corner.
iou
iou(self, other: Self) -> Scalar[dtype]
Calculate Intersection over Union (IoU) with another bounding box.
Args:
- other (
Self): The other bounding box to compare with.
Returns:
Scalar[dtype]: The IoU value, ranging from 0 (no overlap) to 1 (perfect overlap).
intersection_area
intersection_area(self, other: Self) -> Scalar[dtype]
Calculate the area of intersection with another bounding box.
Args:
- other (
Self): The other bounding box to intersect with.
Returns:
Scalar[dtype]: The intersection area, or 0 if boxes don't overlap.
area
area(self) -> Scalar[dtype]
Calculate the area of this bounding box.
Returns:
Scalar[dtype]: The area of the box.
Was this page helpful?
Thank you! We'll create more content like this.
Thank you for helping us improve!