kitcar_ml.utils package

Subpackages

Submodules

kitcar_ml.utils.benchmark module

kitcar_ml.utils.bounding_box module

Classes:

BoundingBox(x1, y1, x2, y2, class_label[, ...])

Class representing a bounding box.

class BoundingBox(x1: float, y1: float, x2: float, y2: float, class_label: str, confidence: Optional[float] = None)[source]

Bases: object

Class representing a bounding box.

The four coordinates are the absolute coordinates in the image.

Attributes:

`x1`	Lower inclusive bound of the bounding box.
`y1`	Left inclusive bound of the bounding box.
`x2`	Upper inclusive bound of the bounding box.
`y2`	Right inclusive bound of the bounding box.
`class_label`	Class that the model gave the bounding box.
`confidence`	Confidence of the model, that this bounding box is correct.
`coordinates`
`center_point`
`width`
`height`
`area`	The area of the bounding box.

Methods:

`shift_and_scale`(shift, scale)	Shift and scale the bounding box :param shift: Vector that translates the bounding box.
`create_bounding_boxes`(boxes, labels[, scores])	Create multiple bounding boxes.
`iou`(box_a, box_b)	Calculate the IOU(Intersection over Union) of two bounding boxes.
`union_area`(box_a, box_b[, intersection_area])	Calculate the union area of two bounding boxes.
`intersection_area`(box_a, box_b)	Calculate the intersection area of two bounding boxes.

x1: float: Lower inclusive bound of the bounding box.

y1: float: Left inclusive bound of the bounding box.

x2: float: Upper inclusive bound of the bounding box.

y2: float: Right inclusive bound of the bounding box.

class_label: str: Class that the model gave the bounding box.

confidence: float = None: Confidence of the model, that this bounding box is correct.

property coordinates

property center_point: Tuple[float, float]

property width: float

property height: float

property area

The area of the bounding box.

(width + 1)x(height + 1), because the boundaries are inside the bounding box.

shift_and_scale(shift, scale)[source]

Shift and scale the bounding box :param shift: Vector that translates the bounding box. :param scale: Vector that increases or decreases the bounding box.

Returns: The scaled and shifted bounding box coordinates.

classmethod create_bounding_boxes(boxes: Iterable[Iterable[int]], labels: Iterable[str], scores: Optional[Iterable[float]] = None)[source]: Create multiple bounding boxes.

classmethod iou(box_a: BoundingBox, box_b: BoundingBox) → float[source]: Calculate the IOU(Intersection over Union) of two bounding boxes.

classmethod union_area(box_a: BoundingBox, box_b: BoundingBox, intersection_area=None) → float[source]: Calculate the union area of two bounding boxes.

static intersection_area(box_a: BoundingBox, box_b: BoundingBox) → float[source]: Calculate the intersection area of two bounding boxes.

kitcar_ml.utils.visualization module

Functions:

`draw_boxes`(image, boxes, labels)	Draw labeled boxes on an image.
`show_labeled_image`(image[, boxes, labels, ...])	Show the image along with the specified boxes around detected objects.

draw_boxes(image, boxes, labels)[source]: Draw labeled boxes on an image.

show_labeled_image(image: Union[Image, ndarray, Tensor], boxes: Optional[Tensor] = None, labels: Optional[List[str]] = None, save_path: Optional[str] = None, show: bool = False)[source]

Show the image along with the specified boxes around detected objects.

Parameters

image – The image that is displayed.
boxes – The bounding boxes on the image.
labels – A list of labels for each bounding box.
save_path – The path to the folder where the image should be saved.
show – If the image should be shown interactively.

Module contents

Utils Package for KITcar-Machine-Learning.