Skip to main content

BEV

Applies Bird's Eye View transform on bounding boxes detected.

This module provides a pipeline element to transform a given frame to Bird's Eye View. This module helps gather more insights of the frame by providing a top-view of the frame.

Examples:

REST API:

WIP

Configuration File:

short-from : BEV:<i-ROI>:<i-dim>

pipeline:
- person-detection-0200:a:0.3:0.3
- BEV:[511,5,1375,3,3,1049,1919,1027]:[1920,1080]
sources:
- address: videos/demo-tracking-28s.mp4
- type: file

BEV Objects

class BEV(Element)

Applies Bird's Eye View transform on bounding boxes detected.

Attributes:

  • name str - The name of the element.
  • ROI int - The Region of Interest where BEV transform will be applied.
  • dim int - The dimensions of the image.

Examples:

ele = BEV("[511,5,1375,3,3,1049,1919,1027]", "[1920,1080]")

extract_params

 | @staticmethod
| def extract_params(model_params)

Extract the parameters for this element.

Arguments:

  • model_params List[str] - The parameters for this element.

Returns:

  • ROI int - The Region of Interest where BEV transform will be applied.
  • dim int - The dimensions of the image.

Raises:

  • ExtractParameterException - If the parameters are not valid.

Examples:

params = BEV.extract_params(["[511,5,1375,3,3,1049,1919,1027]", "[1920,1080]"])
ele = BEV(*params)

process

 | def process(meta)

Process the bounding boxes.

This is the main function for this element. It will take the bounding boxes and a reference plane to transform the frame and bounding boxes to Bird's Eye View.

Arguments:

  • meta dict - The metadata to process.

Returns:

  • meta dict - The processed metadata.

Examples:

ele = BEV("[511,5,1375,3,3,1049,1919,1027]", "[1920,1080]")
meta = ele.process(meta)