Intelligent Vision Control For A Precision Assembly Mobile Robot

Mobile robots are widely used in industry. Computer vision system is an important component in a mobile robot, and it is even more important in a precision assembly mobile robot. Computer vision has been playing a very important role for the robot movement, the robot's arms movement and the corresponding accuracy. An image recognition method with solid accuracy without compromising real-time performance in a specific environment for a mobile robot has been explored for many years.The method provided in this thesis can be easily applied to the actual industry for an assembly mobile robot image recognition technology. Vision SDK and real-time control processing tools (RTX) are used in the image processing. Meanwhile, it is an accuracy and rapid algorithm. This research includes error factor, multilateral scanning, denoising area of the preprocessing and color features in particular circumstances for a mobile robot vision system.Firstly in this thesis images formation, methods and image processing technologies are introduced, and various methods which are currently used, and its application scope, advantages and disadvantages are analysis. Then the methods to improve a robot image recognition technology are discussed. That is, image feature extraction methods which are applied to a precise assembly mobile robot according to work environment features are analysis. Innovative image recognition technologies such as multilateral scanning, pretreatment area and color rapid identification method under a specific condition are proposed .Finally the structure and function are analysis on the robot system design. C + + language and Microsoft Visual Studio compiler are used to completed the precision assembly robot vision system, and fuzzy logic system are used to optimize the robot's path chosen, in order to make a robot have intelligent effectiveness and functionality.