Documentaton - BLOCKS - BLOCK_TEST_SVM

Classify whole images with an SVM.

BK = BLOCK_TEST_SVM() Initializes the block with the default options.

BK = BLOCK_TEST_SVM(BK) Executes the block with options and inputs BK.

Required inputs:

kernel

The kernel computed between the training and test images.

svm

The trained SVM.

Fetchable attributes:

prediction

Returns [PRED, PRED_SEG_IDS, DEC, DEC_LABELS] where PRED are the predictions of the segment labels, PRED_SEG_IDS are the ID of the correspodning segments, DEC is the matrix of the decision functions and DEC_LABELS the corresponding labels. See SVMKERNELTEST()).