add_samples_image_class_knnT_add_samples_image_class_knnAddSamplesImageClassKnnAddSamplesImageClassKnnadd_samples_image_class_knn (Operator)

Name

add_samples_image_class_knnT_add_samples_image_class_knnAddSamplesImageClassKnnAddSamplesImageClassKnnadd_samples_image_class_knn — Add training samples from an image to the training data of a k-Nearest-Neighbor classifier.

Signature

add_samples_image_class_knn(Image, ClassRegions : : KNNHandle : )

Description

add_samples_image_class_knnadd_samples_image_class_knnAddSamplesImageClassKnnAddSamplesImageClassKnnadd_samples_image_class_knn adds training samples from the ImageImageImageimageimage to the k-Nearest-Neighbor (k-NN) given by KNNHandleKNNHandleKNNHandleKNNHandleknnhandle. add_samples_image_class_knnadd_samples_image_class_knnAddSamplesImageClassKnnAddSamplesImageClassKnnadd_samples_image_class_knn is used to store the training samples before a classifier is used for the pixel classification of multichannel images with classify_image_class_knnclassify_image_class_knnClassifyImageClassKnnClassifyImageClassKnnclassify_image_class_knn. add_samples_image_class_knnadd_samples_image_class_knnAddSamplesImageClassKnnAddSamplesImageClassKnnadd_samples_image_class_knn works analogously to add_sample_class_knnadd_sample_class_knnAddSampleClassKnnAddSampleClassKnnadd_sample_class_knn. The ImageImageImageimageimage must have a number of channels equal to NumDimNumDimNumDimnumDimnum_dim, as specified with create_class_knncreate_class_knnCreateClassKnnCreateClassKnncreate_class_knn. ClassRegionsClassRegionsClassRegionsclassRegionsclass_regions must be a tuple containing of at least 2 regions. The order of the regions in ClassRegionsClassRegionsClassRegionsclassRegionsclass_regions determines the class of the pixels. If there are no samples for a particular class in ImageImageImageimageimage an empty region must be passed at the position of the class in ClassRegionsClassRegionsClassRegionsclassRegionsclass_regions. With this mechanism it is possible to use multiple images to add training samples for all relevant classes to the k-NN classifier by calling add_samples_image_class_knnadd_samples_image_class_knnAddSamplesImageClassKnnAddSamplesImageClassKnnadd_samples_image_class_knn multiple times with different images and suitably chosen regions. The regions in ClassRegionsClassRegionsClassRegionsclassRegionsclass_regions should contain representative training samples for the respective classes. Hence, they do not need to cover the entire image. The regions in ClassRegionsClassRegionsClassRegionsclassRegionsclass_regions should not overlap each other, as these samples from overlapping areas would be assigned to multiple classes in the training data, which may lead to a lower classification performance.

Execution Information

Multithreading type: reentrant (runs in parallel with non-exclusive operators).
Multithreading scope: global (may be called from any thread).
Processed without parallelization.

This operator modifies the state of the following input parameter:

KNNHandleKNNHandleKNNHandleKNNHandleknnhandle

During execution of this operator, access to the value of this parameter must be synchronized if it is used across multiple threads.

Parameters

ImageImageImageimageimage (input_object) (multichannel-)image → object (byte / cyclic / direction / int1 / int2 / uint2 / int4 / real)

Training image.

ClassRegionsClassRegionsClassRegionsclassRegionsclass_regions (input_object) region-array → object

Regions of the classes to be trained.

KNNHandleKNNHandleKNNHandleKNNHandleknnhandle (input_control, state is modified) class_knn → (handle)

Handle of the k-NN classifier.

Result

If the parameters are valid, the operator add_samples_image_class_knnadd_samples_image_class_knnAddSamplesImageClassKnnAddSamplesImageClassKnnadd_samples_image_class_knn returns the value 2 ( H_MSG_TRUE) . If necessary an exception is raised.