• documentation
main content

image and video ground truth labeling -凯发k8网页登录

interactive image and video labeling for object detection, semantic segmentation, instance segmentation, and image classification

use the image labeler and video labeler apps to interactively label ground truth data in a collection of images, a video, or a sequence of images. you can label rectangular regions of interest (rois) or polylines for object detection, pixels for semantic segmentation, polygons for instance segmentation, and scenes for image classification. the apps also include computer vision algorithms to automate the labeling of ground truth data for use with detection and tracking algorithms. they also provide an api and workflow that enables you to import your own algorithms to automate the labeling of ground truth data.

apps

image labelerlabel images for computer vision applications
video labelerlabel video for computer vision applications

functions

load data

project cuboids from 3-d world coordinates to 2-d image coordinates
object for storing ground truth data sources
datastore for pixel label data
datastore for bounding box label data
ground truth label data

create label definitions

object for storing, modifying and creating label definitions table
create label definitions table from the label definition creator object
display information about a label, sublabel, or attribute stored in label definition creator object

add

add label to label definition creator object
add sublabel to label in label definition creator object
add attributes to label or sublabel in label definition creator object

edit

modify a label group name
change group name
modify description of label or sublabel in label definition creator object
modify description of attribute in label definition creator object

remove

remove label from label definition creator object
remove sublabel from label in label definition creator object
remove attribute from label or sublabel in label definition creator object

action flags

initialize state for algorithm execution (optional)
run label automation on every frame in interval
terminate automated algorithm (optional)

validation

set up validation (optional)
validate label definition

algorithm

interface for algorithm automation in ground truth labeling
display algorithm settings (optional)
mixin interface for adding temporal context to automation algorithms
mixin interface for performing blocked image automation
implements the blocked image automation algorithm
set reverse algorithm automation flag

store data

ground truth label data
datastore for pixel label data
datastore for bounding box label data
change file paths in ground truth data
write video sequence to video file
time ranges of scene labels from ground truth data
gather label data from ground truth

select labels

select ground truth labels by label group
select ground truth labels by label type
select ground truth labels by label name

enumerate attribute and label types

attribute type enumerations for labeling
label type enumerations for labeling

create training data for object detectors

create training data for an object detector
create training data for semantic segmentation from ground truth
create labeled blockedimage object from set of rois

topics

get started

  • choose an app to label ground truth data
    decide which app to use to label ground truth data: image labeler, video labeler, ground truth labeler, lidar labeler, signal labeler, or medical image labeler.

  • interactively label rectangular rois for object detection, pixels for semantic segmentation, polygons for instance segmentation, and scenes for image classification.

  • interactively label rectangular rois for object detection, pixels for semantic segmentation, polygons for instance segmentation, and scenes for image classification in a video or image sequence.

  • use keyboard shortcuts and mouse actions to increase productivity while using the image labeler app.

  • use keyboard shortcuts and mouse actions to increase productivity while using the video labeler app.

load ground truth data sources


  • create a reader function to load a custom image data source into the ground truth labeler or video labeler.

label ground truth


  • label pixels for training a semantic segmentation network by using a labeling app.

  • label objects using polygons.

  • learn how roi sublabels and attributes work in labeling apps.

  • view a summary of roi and scene labels in a labeling app session.

  • this section describes the differences with using the image labeler app with blocked images.

automate labeling


  • create a custom automation algorithm to use in a labeling app.

  • create a custom automation algorithm function to use in a labeling app.

  • create a time-based custom tracking algorithm to import into a labeling app.

  • create a blocked image custom automation algorithm to import into a labeling app.

use ground truth labels


  • share and store labeled ground truth data exported from labeling apps.

  • learn how the labeling apps store pixel label data.

  • create training data for object detection or semantic segmentation using the image labeler or video labeler.
网站地图