Inference Overlay: Difference between revisions

From Tygron Support wiki
Jump to navigation Jump to search
← Older edit
Maxim@tygron.com (talk | contribs)
Tygron, Tygron_admin, Bureaucrats, Administrators
5,051 edits
Maxim@tygron.com (talk | contribs)
Tygron, Tygron_admin, Bureaucrats, Administrators
5,051 edits
No edit summary
Maxim@tygron.com (talk | contribs)
Tygron, Tygron_admin, Bureaucrats, Administrators
5,051 edits
No edit summary
 
Line 1: Line 1:
[[File:Treesmask.jpg|thumb|right|Trees detected with Inference Overlay.]]
[[File:Treesmask.jpg|thumb|right|Trees detected with Inference Overlay.]]
The AI Inference Overlay is a [[Grid Overlay]] which can spatially identify features using one or more Prequel Grids. Features are identified using a [[Region-based Convolutional Neural Network (Inference Overlay)|Region-based Convolutional Neural Network (RCNN)]] that used a limited amount of proposal regions. This Neural Network takes a subsection (window) of the input grid and either;
The AI Inference Overlay is a [[Grid Overlay]] which can spatially identify features using one or more Prequel Grids. Features are identified using a [[Region-based Convolutional Neural Network (Inference Overlay)|Region-based Convolutional Neural Network (RCNN)]] that used a limited amount of proposal regions. This Neural Network takes a subsection (window) of the input grid and either;
* [[Neural Network (Inference Overlay)#Supported Convolution Types|classifies]] the image as a whole
* [[Region-based Convolutional Neural Network (Inference Overlay)#Supported Convolution Types|classifies]] the image as a whole
* [[Neural Network (Inference Overlay)#Supported Convolution Types|detect one or more features]] in that image.  
* [[Region-based Convolutional Neural Network (Inference Overlay)#Supported Convolution Types|detect one or more features]] in that image.  


It is not possible to run the convolution network on the grid as a whole. Therefore, a marching window strategy is applied instead, where the window is moved over the input grid with a [[Stride fraction (Inference Overlay)|configurable stride]]. This window takes the input values of one or more grids and puts these into tensors representing either floating point values or color channels (red, green and blue). Depending on the dataset used to train the neural network, the tensor values are normalized based on a particular value range. For example, a relative height input tensor for a foliage neural network trained on height data can be normalized using a range of 0 to 40 meters, when 40 meters is considered the largest tree height in the Netherlands.
It is not possible to run the convolution network on the grid as a whole. Therefore, a marching window strategy is applied instead, where the window is moved over the input grid with a [[Stride fraction (Inference Overlay)|configurable stride]]. This window takes the input values of one or more grids and puts these into tensors representing either floating point values or color channels (red, green and blue). Depending on the dataset used to train the neural network, the tensor values are normalized based on a particular value range. For example, a relative height input tensor for a foliage neural network trained on height data can be normalized using a range of 0 to 40 meters, when 40 meters is considered the largest tree height in the Netherlands.

Latest revision as of 15:51, 29 June 2026

Trees detected with Inference Overlay.

The AI Inference Overlay is a Grid Overlay which can spatially identify features using one or more Prequel Grids. Features are identified using a Region-based Convolutional Neural Network (RCNN) that used a limited amount of proposal regions. This Neural Network takes a subsection (window) of the input grid and either;

It is not possible to run the convolution network on the grid as a whole. Therefore, a marching window strategy is applied instead, where the window is moved over the input grid with a configurable stride. This window takes the input values of one or more grids and puts these into tensors representing either floating point values or color channels (red, green and blue). Depending on the dataset used to train the neural network, the tensor values are normalized based on a particular value range. For example, a relative height input tensor for a foliage neural network trained on height data can be normalized using a range of 0 to 40 meters, when 40 meters is considered the largest tree height in the Netherlands.

Foliage Example

Animation of inference with a moving window and Bounding Box detection and a stride fraction of 0.5.

Using a Satellite Overlay of 0.25m detail, foliage features can be identified using an Inference Overlay and enhanced with a Digital Terrain Model Overlay (DTM), a WCS Overlay representing the DSM, Combo Overlay to combine these and an optionally an iterative Max Overlay to enhance the foliage height. For more information, see this how-to.

How-to's

See also

Inference Overlay
Group
Inference
Neural Network • Input tensor • Output tensor
Model attributes
Box overlap • Inference mode • Mask threshold • Prioritize labels • Score threshold • Stride fraction
Prequels
A • B
Result types
Boxes • Labels • Masks • Scores • Debug gt instances • Center points
Neural Networks
Foliage • Solar panels
Overlays
Categorization
Grid • Overlay Type • Parent • Result Child • Result Parent • Result type • Key
Default
City
Administrative
Ownership • Ownership Grid • Zipcodes • Zoning • Neighborhoods • Municipalities • Scenario
Environmental
Heat Stress • Shadow • Traffic Density • Traffic Noise • Traffic NO2 • Livability
Water
Rainfall • Flooding • Groundwater • Watershed
Grid Calculation
Average • Combo • Distance • Sight Distance • Travel Distance • Inference • Iteration
Geo Data
GeoTIFF • WCS • WMS • Image • Area • Attribute • Function
Topography
Satellite • Digital Surface Model • Digital Terrain Model • Building Floors • Underground • Source
Network
Network Distance • Network Overview • Network Ownership


Retrieved from "https://support.tygron.com/w/index.php?title=Inference_Overlay&oldid=87843"