For documentation on the current version, please check Knowledge Base.
This is an old revision of the document!
Image Annotation to Object
This page describes how to use the extension “Image Annotation to Object”
Main Toolbar > Extensions > Image Annotation to Object
The “Image Annotation to Object” extension makes it possible to convert image annotations expressed in pixel coordinates or bounds into 3D coordinates or 3D boxes using pointcloud resource.
Concepts
Image annotations can be defined for mapping run imagery in pixel coordinates and bounds. The image pixels are converted into 3D coordinates or 3D boxes in a vector file using the mapping run pointcloud.
The 2D annotations and tags are overlayed on the opened image.
Source
Pixel Definitions
Select how the annotations are defined:
- Coordinates: every annotation is defined by 1 coordinate
- Bounds: every annotation is defined by 2 pixel coordinates that form the bounds of the rectangle
Reference Files
Choose how the reference files are structured:
- One file for every image
- One file for all images
Supported formats are .csv and .xml.
Target
After selecting the target file location and CRS, choose to create 3D point objects or boxes.
Parameters
Advanced Parameters
Search Slice Thickness
The thickness of the slices created from camera position to pixel point.
For creating 3D Coordinates:
If a slice has an object, don't search in the following slices.
For creating 3D Boxes:
Do search in the following slices up until the Search Dist Max is reached.
Search Dist Max
The maximum distance from the camera to the detected object.
Cluster Pts Min
The minimum number of points to have a cluster.
Search Radius (Box from Pix Coord)
The search radius around the direction camera to pixel position to search for the 3D object.
Ground margin
The ground margin used to identify a 3D box reached the ground to be able to end the clustering.
Voxel Size
The voxel size used to cluster objects.
Voxel Pts Min
The minimum number of points to have a voxel.