Runs a trained neural network model on an image. This function allows you to use artificial intelligence to detect objects in an image.
Caution: This function is an advanced function. To use this function, it is recommended to read the documentation of the neural network used (specifics, expected results and interpretation).
// 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)
Syntax
<Result> = AIDetectModel(<Model> , <Image>)
<Result>: Array of array
Array of matrices containing the result of the execution of the model.
This array is specific to each template and must be known by the developer.
<Model>: aiNeuralNetworkModel variable
Name of the aiNeuralNetworkModel variable describing the characteristics of the neural network used.
Caution: The different characteristics of this Variable are specific to each template and must be known by the developer.
<Image>: String, Image or Image control
Image to be analyzed. The image can correspond to: - a variable of type Image,
- the name and path of the image,
- the name and path of a PDF file,
- an Image memo item,
- an Image control.
Remarks
- The AI engine used is OpenCV. This engine reads the AI models and executes them.
- The supported neural network models are:
- caffe,
- tensorflow,
- darknet,
New in version 28onnx
- The model must be trained (weights are known).
- The expected extensions according to the models are:
- Configuration:
- Caffe: *.prototxt
- Tensorflow: *.pbtxt
- Darknet: *.cfg
- Poids:
- Caffe: *.caffemodel
- Tensorflow: *.pb
- Darknet: *.weights
New in version 28Open Neural Network Exchange (ONNX): *.onnx
Business / UI classification: Business Logic