Research On Visual Perception And Recovery Algorithm Of Intelligent Magic Cube Robot

Recently, the development of intelligent robot technology receives attention in the field of high-tech. Robots has been increasingly applied in people’s life and production. It plays an important role in our life.Intelligent Cube Robot is a robot control system which can reset the messy cube by any way. It involves some technology, such as Digital Image Processing, Computer Vision, Pattern Recognition, Artificial Intelligence etc. The thesis researched mainly the image of cube and implement of the algorithm which to reset cube. The thesis has builded a virtual interacted system which could reset the cube automatically. It also has constrcted LEGO Cube Robot by LEGO components. The purpose is to test and verify the accuracy and effectiveness of system which designed.The content of this thesis is mainly divided into two parts:the first is to identify the position and color of the small pieces of Magic Cube based on feature information of analysis of Cube image. It mainly includes Cube image preprocessing, segmentation, targeting, color recognition and so on. This thesis proposes a new application methods based on the previous theory for research object. The second part is making an in-depth analysis on structure, properties recovery algorithm of Magic Cube. The rotation of the Magic Cube is descripted by mathematical methods. The computer solution method of Magic Cube has been given in the variety of states. The third part is the combination of recovery algorithm of Magic Cube and digital image processing technology that makes the intelligent robot reset the Magic Cube based on visual image processing in the virtual system. Then the correct the recovervy algorithm is examined and verified by the recovery operation of LEGO Magic Cube Robot.This thesis is a comprehensive study based on the visual image of theory knowledge combined with the Magic cube educational function. A lot of experiments on Magic Cube recovery have been done in this thesis. The test results of the color recognition rate and the recovery rate of the system has been analyzed and discussed. The rapidity and accuracy of the system has been analyzed and discussed. Finally, these experiments show that the design and the algorithm of this thesis are reasonable and feasible.