Is is possible to create a hyperplane in 3D for SVM? I have 3 features to train the multi-class svm

Question

enter image description here

channel 1 2 3 represent 3 features. The label is the target values (integers from 0 to 17). Thus I have 18 classes to classify my data points to. I was imagining a 3D system where x y z are the three features and got a decision boundary surface in 3D. I used a rbf kernel for training.