eng Science and Education Publishing American Journal of Mechanical Engineering 2014-10-15 2 7 216 218 10.12691/ajme-2-7-9 AJME2014279 article Pavol Bezák 1 Yury Rafailovich Nikitin 1 Pavol Bo?ek pavol.bozek@stuba.sk 2 Institute of Applied Informatics, Automation and Mathematics, Faculty of Materials Science and Technology, Slovak University of Technology, Trnava, Slovakia Kalashnikov Izhevsk State Technical University, Mechatronic Systems Department, Izhevsk, Russia Object grasping by robot hands is challenging due to the hand and object modeling uncertainties, unknown contact type and object stiffness properties. To overcome these challenges, the essential purpose is to achieve the mathematical model of the robot hand, model the object and the contact between the object and the hand. In this paper, an intelligent hand-object contact model is developed for a coupled system assuming that the object properties are known. The control is simulated in the Matlab Simulink/ SimMechanics, Neural Network Toolbox and Computer Vision System Toolbox.. http://pubs.sciepub.com/ajme/2/7/9/ajme-2-7-9.pdf robot hand modeling grasping convolutional neural networks deep learning object recognition pose estimation