Tesi di laurea specialistica in Ingegneria Informatica presso il Politecnico di Milano
- Title:
Video recognition of human activities for use in humanoid robotics
- Author:
Giovanni Paolo Barni & Decio Denis Bernardo
- Description:
The thesis analyses and develops the theories and methods used in the systems of
imitation learning in robotic applications and uses software for teaching and recognizing human movements. The programme is based on the theory that nine different movements of the human arm could be distinguished when placed in front of the visual learning system. This then communicated the coordinate in the space of the joints to the robotic simulator or the robot MAXIMUM_ONE (a prototype developed in the AirLab department of Milan Polytechnic). The system uses a variation of the CAMSHIFT algorithm to teach the
movements of the hands by video and extending the use to everything whose movement you wish to study. MATLAB and other vision toolboxes such as Image Acquisition Toolbox and Image Processing Toolbox, and robotic toolboxes such as Robotics Toolbox were used to develop this thesis.
The system is oriented to the various state (that is known position) by using statistical algorithms and operating on images to obtain the best results.
The entire project is oriented towards the study of human activities using video and for robotics, low-level hardware systems: Intel Celeron 1,8 MHz and a standard webcam.
Good stuff
amightyo 5 months ago
Finalmente il lavoro di una vita realizzato e pubblicato!
-LAUREA SPECIALISTICA FACOLTA' ING. INFORMATICA-
dextyle 3 years ago
ke figata prof!!!!!
martypiro 3 years ago