With the continuous development of information technology,brush calligraphy as a traditional Chinese art is in urgent need of inheritance and promotion.More and more humanoid robot calligraphy has caught the public's eye,Humanoid robot has become an important research direction in the field of service robot.Taking Nao robot as the research object,this paper studies the robot's soft pen calligraphy from three aspects: robot kinematics,Chinese character information extraction and writing control methods.The main contents of this paper are as follows:1.The D-H parameter modeling method is used to model the right arm of the NAO robot,and then the forward and inverse kinematics analysis is carried out to find out the position and posture matrix of the right arm when holding the pen and the solution of each joint angle.The simulation of the right arm holding the pen is verified by using the robotics toolbox in MATLAB software,and the inverse kinematics analysis is carried out at the same time.Monte Carlo method is used to solve the working space of the right arm of the NAO robot.The point cloud map of the right arm's working space and the point cloud map of the right arm's writing desk space are obtained through MATLAB simulation,and the writing position and area size of a single Chinese character are determined.2.The joint space and Cartesian space of the NAO robot are planned respectively.The trajectory planning of the right arm of the NAO robot in the Cartesian coordinate system in the workspace is discussed in detail.The position and attitude interpolation methods of two kinds of curve motion,straight line and arc,are described and verified by MATLAB software simulation.The attitude interpolation is analyzed by the unit quaternion algorithm.The motion process of the two curves is divided into five parts: acceleration,deceleration,uniform speed,acceleration and deceleration.The acceleration and deceleration are planned by the five part S-shaped speed curve control algorithm to make the interpolation motion smooth and continuous.3.This paper studies the structure of Chinese characters and the characteristics of regular script strokes,preprocesses the characters and pictures recognized by Nao robot,smoothes and filters the images by median filter,binarizes them by Otsu,prepares for thinning,extracts the skeleton information of Chinese characters by Hilditch thinning algorithm,and carries out skeleton processing based on the improved maximum circle algorithm Through the optimization algorithm and Freeman direction chain code method,the information of stroke width and stroke order is extracted,and the stable and clear information of Chinese characters is obtained.4.The writing behavior and characteristics of the brush are analyzed,and a control method of robot brush writing is proposed.At last,the experimental results show that the writing effect is better by comparing with the traditional Nao robot.The research scheme of this paper is expected to be extended to various robots to write different fonts. |