Inference Overlay - Revision history

Maxim@tygron.com at 15:51, 29 June 2026

2026-06-29T15:51:44Z

← Older revision		Revision as of 15:51, 29 June 2026
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.

Maxim@tygron.com at 15:48, 29 June 2026

2026-06-29T15:48:50Z

← Older revision		Revision as of 15:48, 29 June 2026
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 ~~Convolution~~ [[Neural Network]] that used a limited amount of proposal regions ~~(RCNN)~~. 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		* [[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.		* [[Neural Network (Inference Overlay)#Supported Convolution Types\|detect one or more features]] in that image.

Rudolf@tygron.nl at 11:05, 24 June 2026

2026-06-24T11:05:17Z

← Older revision		Revision as of 11:05, 24 June 2026
Line 33:		Line 33:
	* [[How to train your own AI model for an Inference Overlay]]		* [[How to train your own AI model for an Inference Overlay]]
	* [[How to select specific input data for AI Inference]]		* [[How to select specific input data for AI Inference]]
	* [[How to create AI ~~train~~ data with QGIS]]		* [[How to create AI training data with QGIS]]
	}}		}}
	{{InferenceOverlay nav}}		{{InferenceOverlay nav}}
	{{Overlay nav}}		{{Overlay nav}}

Maxim@tygron.com: /* Foliage Example */

2026-06-16T09:10:50Z

Foliage Example

← Older revision		Revision as of 09:10, 16 June 2026
Line 8:		Line 8:
	==Foliage Example==		==Foliage Example==
	[[File:inference_foliage_animated.gif\|frame\|right\|Animation of inference with a moving window and Bounding Box detection and a [[Stride fraction (Inference Overlay)\|stride fraction]] of 0.5.]]		[[File:inference_foliage_animated.gif\|frame\|right\|Animation of inference with a moving window and Bounding Box detection and a [[Stride fraction (Inference Overlay)\|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\|WCS Overlay]] representing the [[DSM]], [[Combo Overlay]] to combine these and an optionally an iterative [[Average ~~(overlay)~~\|Max Overlay]] to enhance the foliage height. For more information, see this [[How to detect foliage using an Inference Overlay\|how-to]].		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\|WCS Overlay]] representing the [[DSM]], [[Combo Overlay]] to combine these and an optionally an iterative [[Average Overlay\|Max Overlay]] to enhance the foliage height. For more information, see this [[How to detect foliage using an Inference Overlay\|how-to]].
	<gallery widths="200">		<gallery widths="200">
	foliage_inference_labels.jpg\|[[Labels result type (Inference Overlay)\|Labeled features]] on 0.25m satellite image		foliage_inference_labels.jpg\|[[Labels result type (Inference Overlay)\|Labeled features]] on 0.25m satellite image

Maxim@tygron.com at 09:10, 16 June 2026

2026-06-16T09:10:18Z

← Older revision		Revision as of 09:10, 16 June 2026
Line 8:		Line 8:
	==Foliage Example==		==Foliage Example==
	[[File:inference_foliage_animated.gif\|frame\|right\|Animation of inference with a moving window and Bounding Box detection and a [[Stride fraction (Inference Overlay)\|stride fraction]] of 0.5.]]		[[File:inference_foliage_animated.gif\|frame\|right\|Animation of inference with a moving window and Bounding Box detection and a [[Stride fraction (Inference Overlay)\|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\|WCS Overlay]] representing the [[DSM]], [[Combo Overlay]] to combine these and an optionally an iterative [[~~Avg & interpolation~~ (overlay)\|Max Overlay]] to enhance the foliage height. For more information, see this [[How to detect foliage using an Inference Overlay\|how-to]].		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\|WCS Overlay]] representing the [[DSM]], [[Combo Overlay]] to combine these and an optionally an iterative [[Average (overlay)\|Max Overlay]] to enhance the foliage height. For more information, see this [[How to detect foliage using an Inference Overlay\|how-to]].
	<gallery widths="200">		<gallery widths="200">
	foliage_inference_labels.jpg\|[[Labels result type (Inference Overlay)\|Labeled features]] on 0.25m satellite image		foliage_inference_labels.jpg\|[[Labels result type (Inference Overlay)\|Labeled features]] on 0.25m satellite image

Maxim@tygron.com: Undo revision 63569 by Maxim@tygron.com (talk)

2025-10-17T15:09:44Z

Undo revision 63569 by Maxim@tygron.com (talk)

← Older revision		Revision as of 15:09, 17 October 2025
Line 32:		Line 32:
	* [[How to create the input overlay for leafless trees detection using the Inference Overlay]]		* [[How to create the input overlay for leafless trees detection using the Inference Overlay]]
	* [[How to train your own AI model for an Inference Overlay]]		* [[How to train your own AI model for an Inference Overlay]]
	* [[How to ~~detect Trees in private yards using an~~ Inference ~~Overlay~~]]		* [[How to select specific input data for AI Inference]]
	* [[How to create AI train data with QGIS]]		* [[How to create AI train data with QGIS]]
	}}		}}
	{{InferenceOverlay nav}}		{{InferenceOverlay nav}}
	{{Overlay nav}}		{{Overlay nav}}

Maxim@tygron.com at 15:08, 17 October 2025

2025-10-17T15:08:34Z

← Older revision		Revision as of 15:08, 17 October 2025
Line 32:		Line 32:
	* [[How to create the input overlay for leafless trees detection using the Inference Overlay]]		* [[How to create the input overlay for leafless trees detection using the Inference Overlay]]
	* [[How to train your own AI model for an Inference Overlay]]		* [[How to train your own AI model for an Inference Overlay]]
	* [[How to ~~select specific input data for AI~~ Inference]]		* [[How to detect Trees in private yards using an Inference Overlay]]
	* [[How to create AI train data with QGIS]]		* [[How to create AI train data with QGIS]]
	}}		}}
	{{InferenceOverlay nav}}		{{InferenceOverlay nav}}
	{{Overlay nav}}		{{Overlay nav}}

Maxim@tygron.com: Reverted edits by Maxim@tygron.com (talk) to last revision by Frank@tygron.nl

2025-10-17T15:07:46Z

Reverted edits by Maxim@tygron.com (talk) to last revision by Frank@tygron.nl

← Older revision		Revision as of 15:07, 17 October 2025
Line 5:		Line 5:

	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.

	~~The {{software}} contains several trained models, such as the foliage example explained below. However it is also possible to train your own model with the [[AI Suite]].~~

	==Foliage Example==		==Foliage Example==
Line 22:		Line 20:
	{{article end		{{article end
	\|seealso=		\|seealso=
	* [[AI Suite\|AI Suite (for more information on training your own model)]]
	* [[Model attributes (Inference Overlay)]]		* [[Model attributes (Inference Overlay)]]
	* [[Neural Network]]		* [[Neural Network]]
Line 33:		Line 30:
	* [[How to import trees based on an Inference Overlay]]		* [[How to import trees based on an Inference Overlay]]
	* [[How to detect solar panels using an Inference Overlay]]		* [[How to detect solar panels using an Inference Overlay]]
	* [[How to ~~detect Trees in private yards~~ using an Inference Overlay]]		* [[How to create the input overlay for leafless trees detection using the Inference Overlay]]
			* [[How to train your own AI model for an Inference Overlay]]
	* [[How to select specific input data for AI Inference]]		* [[How to select specific input data for AI Inference]]
	* [[How to ~~evaluate an~~ AI ~~model~~]]		* [[How to create AI train data with QGIS]]
	}}		}}
	{{InferenceOverlay nav}}		{{InferenceOverlay nav}}
	{{Overlay nav}}		{{Overlay nav}}

Maxim@tygron.com at 15:05, 17 October 2025

2025-10-17T15:05:20Z

← Older revision		Revision as of 15:05, 17 October 2025
Line 33:		Line 33:
	* [[How to import trees based on an Inference Overlay]]		* [[How to import trees based on an Inference Overlay]]
	* [[How to detect solar panels using an Inference Overlay]]		* [[How to detect solar panels using an Inference Overlay]]
	* [[How to ~~create the input overlay for leafless trees detection~~ using ~~the~~ Inference Overlay]]		* [[How to detect Trees in private yards using an Inference Overlay]]
	* [[How to select specific input data for AI Inference]]		* [[How to select specific input data for AI Inference]]
	* [[How to evaluate an AI model]]		* [[How to evaluate an AI model]]

Maxim@tygron.com at 10:33, 16 October 2025

2025-10-16T10:33:11Z

← Older revision		Revision as of 10:33, 16 October 2025
Line 5:		Line 5:

	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.

			The {{software}} contains several trained models, such as the foliage example explained below. However it is also possible to train your own model with the [[AI Suite]].

	==Foliage Example==		==Foliage Example==

← Older revision		Revision as of 15:51, 29 June 2026
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.

← Older revision		Revision as of 15:48, 29 June 2026
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 ~~Convolution~~ [[Neural Network]] that used a limited amount of proposal regions ~~(RCNN)~~. 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		* [[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.		* [[Neural Network (Inference Overlay)#Supported Convolution Types\|detect one or more features]] in that image.