LIS Classification - point cloud classification and feature extraction
This package contains tools for supervised and unsupervised point cloud classification, including the derivation of geometric primitives like lines, planes and volumes, and the classification of point clouds into the classes ground, building, vegetation, power lines and pylons.
Tool: Attach Classification to Point Cloud
| Features |
|---|
| • attaching the result of a supervised classification back to the point cloud |
| Applications |
|---|
| • supervised point cloud classification |
Tool: Classify Point Cloud by Polygon
| Features |
|---|
| • assignment of classification identifiers from polygons to points based on point-in-polygon checks |
| • additional identifier for undefined points |
| • filter options to process only certain classes |
| • optional polygon buffering |
| Applications |
|---|
| • point cloud classification |
Tool: Clean Building Facades
| Features |
|---|
| • removal of (misclassified) vegetation points from building facades |
| • optional: minimum height above ground filter |
| • used-defined target class |
| Applications |
|---|
| • point cloud classification |
Tool: Clean Building Roofs
| Features |
|---|
| • removal of (misclassified) vegetation points from building roofs (e.g. antennas, chimneys) |
| • used-defined target class |
| Applications |
|---|
| • point cloud classification |
Tool: Create Classification Error Matrix
| Features |
|---|
| • evaluate a classification result with a reference classification |
| • calculate confusion matrix as well as user, producer and overall accuracy |
| Applications |
|---|
| • evaluation of classification results with a reference classification |
Tool: Create Feature Grid Stack
| Features |
|---|
| • creation of a grid stack with point cloud features for supervised classification |
| Applications |
|---|
| • supervised point cloud classification |
Tool: Create Training Areas from Table
| Features |
|---|
| • creation of training areas from a table with segment ids and classification ids |
| Applications |
|---|
| • supervised point cloud classification |
Tool: Eigenvalue Classification
| Features |
|---|
| • template matching of eigenvalue patterns for feature classification in point clouds |
| • pattern templates can be reviewed and modified |
| • optional output of classification quality |
| • pre-defined pattern templates: isolated point, end of line, line, plane, half plane, quarter plane, two planes, three planes, two planes (30 degree), volume |
| • optional: classification of lines and planes in horizontal, intermediate and vertical |
| Applications |
|---|
| • feature extraction |
| • point cloud classification |
Tool: Enhanced Point Cloud Classification
| Features |
|---|
| • classification of building and vegetation points |
| • input: point cloud with ground classification, height above ground and segmentation result |
| • two options for building classification: based on point cloud features (optimized for LiDAR) and based on surface roughness (optimized for photogrammetric point clouds) |
| • height slicing of vegetation points into three classes |
| Applications |
|---|
| • point cloud classification |
Tool: Geometric Primitives Classification
| Features |
|---|
| • classification of geometric primitives derived by Eigenvalue classification |
| • output classes: undefined, ground, ground inventory, wall, roof, vertical pole, horizontal pole |
| Applications |
|---|
| • feature extraction |
| • point cloud classification |
Tool: Ground Classification
| Features |
|---|
| • classification of ground points by progressive TIN densification |
| • optional: additional use of pre-computed segmentation result |
| • optional: seed filtering (several criteria) |
| • optional: height above ground computation |
| • optional: output of final TIN as 3D polygon shapes layer |
| Applications |
|---|
| • point cloud classification |
Tool: Ground Classification and Basic Filtering [muli-tile]
| Features |
|---|
| • classification of ground points by progressive TIN densification |
| • works directly on virtual point cloud datasets (multi-tile) |
| Applications |
|---|
| • point cloud classification (multi-tile) |
| • mass data processing |
Tool: Power Line Classification
| Features |
|---|
| • detection and classification of power line cables |
| • optional: classification of vertical and elongated objects as poles |
| • optional: filtering with attribute threshold |
| Applications |
|---|
| • point cloud classification |
Tool: Power Line Refinement
| Features |
|---|
| • refinement of power line classification by filter options |
| • separation of individual cables by segmentation and region growing |
| Applications |
|---|
| • point cloud classification |
| • point cloud segmentation |
Tool: Pylon Classification
| Features |
|---|
| • classification of pylons |
| • input: point cloud with (refined) power line classification |
| Applications |
|---|
| • point cloud classification |