Methods for sign language dactylemes alphabet classification: 
band separation and support vector machines (SVM) are investigated. 
The different methods of features formation are proposed. 
The advantages and disadvantages of each of the proposed methods are discussed.