Synthesis of a man machine interface environment for object recognition using tactile data

Presents a Man Machine Interface Environment based on a blackboard structure that has been developed to perform object recognition using tactile data. The structure of the system is divided into several parts, each one performs a task and allows the whole system to cope with: the robot control and the data acquisition, the reference library builder, the rules producer for recognition algorithm, and the pattern recognition. The approach taken relies on a graphics display and the use of windows that are associated with dialog boxes. The achievement of this software is to integrate two expert systems, the one which has to deal with the recognition task, plus a second one, which will manage all the packages described and will provide to the user some help when he is incorporating new objects in the reference library, such as when the user is building the rules, linked to this object for the matching expert system