add_samples_image_class_mlp — Add training samples from an image to the training data of a
add_samples_image_class_mlp adds training samples from the
Image to the multilayer perceptron (MLP) given by
add_samples_image_class_mlp is used to
store the training samples before a classifier to be used for the
pixel classification of multichannel images with
classify_image_class_mlp is trained.
add_samples_image_class_mlp works analogously to
add_sample_class_mlp. Because here the MLP is always used
OutputFunction = 'softmax'
must be specified when the MLP is created with
create_class_mlp. The image
Image must have a
number of channels equal to
NumInput, as specified with
create_class_mlp. The training regions for the
NumOutput pixel classes are passed in
ClassRegions must be a tuple
NumOutput regions. The order of the regions in
ClassRegions determines the class of the pixels. If there
are no samples for a particular class in
Image an empty
region must be passed at the position of the class in
ClassRegions. With this mechanism it is possible to use
multiple images to add training samples for all relevant classes to
the MLP by calling
times with the different images and suitably chosen regions. The
ClassRegions should contain representative
training samples for the respective classes. Hence, they need not
cover the entire image. The regions in
not overlap each other, because this would lead to the fact that in
the training data the samples from the overlapping areas would be
assigned to multiple classes, which may lead to slower convergence
of the training and a lower classification performance.
This operator modifies the state of the following input parameter:
During execution of this operator, access to the value of this parameter must be synchronized if it is used across multiple threads.
→object (byte / cyclic / direction / int1 / int2 / uint2 / int4 / real)
Regions of the classes to be trained.
MLPHandle(input_control, state is modified) class_mlp
If the parameters are valid, the operator
add_samples_image_class_mlp returns the value TRUE. If
necessary an exception is raised.