Overview

Blurs and redacts any detected objects or regions in the video

Inputs & Outputs

Inputs : 1, Media Format : Raw Video
Outputs : 1, Media Format: Raw Video
Output Metadata : None

Properties

Property	Description	Type	Default	Required
`enabled`	Enabled	`bool`	`true`	No
`roi_labels`	Regions of interest labels	`string`	—	No
`rois`	Regions of interest to blur and redact. Conditional on `roi_labels`. Format: comma-separated normalized `x,y` coordinate pairs; separate multiple polygons with semicolons (for example, `0.1,0.1,0.9,0.1,0.9,0.9`).	`string`	`null`	No
`object_labels`	ex. `car`,`person`,`face.mask`. Process detected objects with specified types depending on the mode. Use the `object_label.class_label` format to blur objects that also have a specific class.	model-labels	`null`	No
`mode`	Include or exclude objects that match the specified types. Options: Include (`include`): Include objects that match the specified object types; Exclude (`exclude`): Exclude objects that don't match the specified object types.	`enum`	`include`	No
`blur_type`	Blur method to use. Options: Simple (`simple`); Dense (`gaussian`); Opaque (`opaque`); Pixelated (`pixel`).	`enum`	`pixel`	No
`intensity`	Blur intensity. Range: minimum `0`, maximum `10`.	`integer`	`7`	No

Output Metadata

Node does not insert any output metadata; upstream metadata continues through the pipeline

Customize

In order to customize this node for your use case, use the code below inside a Function Node.

from lumeopipeline import VideoFrame 
import cv2                   

def process_frame(frame: VideoFrame, object_labels = None, **kwargs) -> bool:
  width = frame.video_info().width
  height = frame.video_info().height

  with frame.data() as mat:
    meta = frame.meta()
    if meta is not None:
      objects = meta.get_field("objects")
      if objects is not None:
        object_label_list = object_labels.lower().replace(" ","").split(",") if object_labels is not None else []
          
        for object in objects:
          if object['label'] in object_label_list or len(object_label_list) == 0:
            object_rect = object['rect']
            object_rect = {key: int(value) for key, value in object_rect.items()}
            y = object_rect['top']
            yEnd = object_rect['top'] + object_rect['height']
            x = object_rect['left']
            xEnd = object_rect['left'] + object_rect['width']
            mat[y:yEnd,x:xEnd] = cv2.blur(mat[y:yEnd,x:xEnd], (23,23) )
              
  return True