ONLINE HELP
 WINDEVWEBDEV AND WINDEV MOBILE

This content has been translated automatically.  Click here  to view the French version.
Help / WLanguage / WLanguage functions / Standard functions / Image detection functions
  • Properties specific to aiNeuralNetworkModel variables
WINDEV
WindowsLinuxUniversal Windows 10 AppJavaReports and QueriesUser code (UMC)
WEBDEV
WindowsLinuxPHPWEBDEV - Browser code
WINDEV Mobile
AndroidAndroid Widget iPhone/iPadIOS WidgetApple WatchMac CatalystUniversal Windows 10 App
Others
Stored procedures
aiNeuralNetworkModel (Variable type)
In french: iaModèleRéseauDeNeurones
The aiNeuralNetworkModel type is used to define all the advanced characteristics of a neural network used by AIDetectModel. You can define and change the characteristics of this neural network using different WLanguage properties.
Remark: For more details on the declaration of this type of variable and the use of WLanguage properties, see Declaring a variable.
Example
// Déclaration de l'image
MonImage is Image
MonImage = IMG_Test

// Déclarer un modèle
MonIAModèleRN is aiNeuralNetworkModel

MonIAModèleRN.Configuration = "MonModel.cfg"
MonIAModèleRN.TrainedWeights = "MonModel.weights.pb"
MonIAModèleRN.PixelScaleFactor = 1.0

// Attention: les dimensions X et Y dépendent du modèle.
// Si les dimensions spécifiées ne correspondent pas au modèle,
// la fonction IAModèleDétecte renverra une erreur.
MonIAModèleRN.XDimension = 300 
MonIAModèleRN.YDimension = 300 

MonIAModèleRN.AverageIntensityR = 104
MonIAModèleRN.AverageIntensityG = 117
MonIAModèleRN.AverageIntensityB = 113
MonIAModèleRN.RGBColor = True

montabMatrice is array of 1 array of 1 by 1 by 200 by 7 reals

// Exécuter le modèle
montabMatrice = AIDetectModel(MonIAModèleRN, MonImage)
Remarks

Properties specific to aiNeuralNetworkModel variables

The following properties can be used to handle a neural network model:
Property nameType usedEffect
AverageIntensityBIntegerAverage intensity of Color Blue (B) training data.
This property is optional.
AverageIntensityGIntegerAverage intensity of Color Green (V) training data.
This property is optional.
AverageIntensityRIntegerAverage intensity of Color Red (R) training data.
This property is optional.
ConfigurationCharacter stringFull path of the file that contains the configuration of the model.
Remark: The possible types of configurations are:
  • Caffe: *.prototxt
  • Tensorflow: *.pbtxt
  • Darknet: *.cfg
OutputLayerNameCharacter stringName of the output neural network layer.
PixelScaleFactorRealUsed to scale pixel values. Set to 1.0 by default (no scaling).
New in version 2024
RequiresTransposition
Boolean
  • if transposition is required,
  • False otherwise.
When to use Property
The OpenCV engine creates matrices from images by defining them as "DimensionX, DimensionY, Number of chanel"..
Some models expect to have a "Nombre de chanel, DimensionX, DimensionY" matrix.. We then need to modify the Image matrix representation in OpenCV.
RGBColorBoolean
  • True if the provided images are encoded in RGB. In this case, the conversion to BGR encoding will be done automatically.
  • False (default value) if the provided images are encoded in BGR.
This property is optional.
TrainedWeightsCharacter string Full path of the file that contains the trained weights of the model.
Remark: The possible types of weights are:
  • Caffe: *.caffemodel
  • Tensorflow: *.pb
  • Darknet: *.weights
  • Open Neural Network Exchange (ONNX): *.onnx
XDimensionIntegerImage size required by the model: X value.
By default, this property corresponds to the width of the image.
YDimensionIntegerImage size required by the model: Y value.
By default, this property corresponds to the height of the image.
Minimum version required
  • Version 27
This page is also available for…
Comments
Click [Add] to post a comment

Last update: 01/12/2024

Send a report | Local help