Instruments and Systems: Monitoring, Control, and Diagnostics Annotation << Back The use of active appearance models for the design of non-contact interface for control of mechatronic systems MOROZOVA T.Yu. AKIMOV D.A. TEREKHIN I.V. The article is devoted to methods and means of addressing the problem of creating a contactless control of mechatronic systems. Considered Bayesian networks, support vector, tools and algorithms for neural network technology, the classifier Viola-Jones, deformable model. The methods and tools based on pattern recognition tools. The task of face detection is considered as a special case of the problem of recognition. Feature vector used for classification of images into two classes. The main attention is paid to active appearance models, which are more effective in the design of a contactless interface. This model allows us to simulate two types of parameters: the parameters associated with the form, and the parameters associated with the statistical model of the image. The use of intelligent technology allows the model to study the set pre-marked images. Layout image using the algorithm for determining the labels. The use of the active appearance model allows to minimize the error of identification, to optimize the computational process, thus providing the program in real time. Considered tools for intelligent contactless interface. For the practical implementation of the developed algorithms learning and adapting the appearance of active models is proposed to use specialized software library AAM Toolbox. Keywords: non-contact interface, the active appearance model, deformable model method labels, training sample, library AAM Toolbox. Contacts: E-mail: TMorozova2006@rambler.ru E-mail: igor.terekhin @ gmail.com E-mail: akimovdmitri@gmail.com Pp. 02-12.