Dissertation > Excellent graduate degree dissertation topics show

Research on Robot Uncalibrated Visual Servoing Key Technique

Author: LiMu
Tutor: ZhaoJie
School: Harbin Institute of Technology
Course: Mechanical and Electronic Engineering
Keywords: Uncalibrated visual servo Information Fusion Adaptive Computing dual threshold Dynamic residuals estimated Jacobian matrix changes estimated Robot
CLC: TP242.2
Type: PhD thesis
Year: 2008
Downloads: 629
Quote: 3
Read: Download Dissertation

Abstract


In recent years, robotics has become one of the representative strategic high-tech fields, it makes the traditional industrial mode of production is undergoing a fundamental change to have a profound impact on the development of human society. With the rapid development of computer vision and computer hardware technology, the visual information with the robot control combine to form a visual servo system to make robots capable of intelligent interaction with the external environment, is today one of the main directions of development of the robot. It can be predicted, intelligent robot with visual will be more widely used. The traditional robot visual servo control system is based on calibration technique, the entire servo system control accuracy is largely dependent on the accuracy of the calibration. However, in practice, due to various reasons, such visual servo method based on the calibration of the robot has been greatly limited. Uncalibrated visual servo started to become a hot research topic in the field of robot visual servo control, the so-called \driver robot motion, system errors converge to a permissible error. This article is for the \Calibration target in the visual servo feature points extraction method. Region color and edge information fusion method, the image is relatively stable segmentation, and then extract the centroid. The method takes full advantage of the characteristics of wavelet multi-resolution analysis, image edge detection. The application of traditional wavelet image edge detection, the threshold needed to filter the candidate edge point mode non-maxima suppression strike edge. Threshold strike with people experience artificially setting requires repeated trial and error comparison in order to arrive at a final results, in addition, this single threshold automatic strike method also can not achieve the accurate edge detection, these deficiencies limit the wavelet Edge detection technology in the actual application. To solve this problem, the paper presents adaptive strike a double threshold algorithm based on class variance minimization principle, no artificial set any coefficients and parameters. This adaptive threshold method based on gradient edge detection technique is equally applicable. Using self-selected area color segmentation image segmentation and edge detection information fusion technology, to achieve a more stable segmentation, more accurate extraction of the target centroid, shall be free hopping phenomenon color segmentation and edge detection can not identify the interested target problems. This paper presents a dynamic uncalibrated visual servo control method, the system dynamic residuals estimated. Uncalibrated visual servo control technology or only for a static target or dynamic target, but can not get rid of affect system dynamics residuals. Therefore, this paper, based on nonlinear minimum variance method to control the robot to track moving targets, the use of dynamic quasi-Newton method estimated image Jacobian matrix, iterative least-squares method to improve the stability of the system proposed estimated the dynamic residuals of the dynamic system method, robot tracking of moving targets. Uncalibrated visual servo control structure dynamic visual servo control algorithm can be applied to the \\Propose a method to estimate the change of each time increment time image Jacobian matrix. By nonlinear objective function to minimize visual information tracking dynamic image. Finally, the use of related hardware set up a set of uncalibrated visual servo experimental system. Through multiple sets of experimental measurement of the performance of the system, and achieved the desired results, and to verify the effectiveness of the algorithm.

Related Dissertations

  1. The Design of a DSP-Based Robot Speech Command Recgnition System,TN912.34
  2. Tongue Feature Extraction and Research of Fusion Classification,TP391.41
  3. Development of Experimental Robot Control System Based on ARM,TP242.6
  4. Multi-Sensor Information Fusion and Its Applications on Wearable Computer,TP202
  5. Design and Realize of Family Cleaning Robot Path-Coverage System,TP242
  6. Humanoid Robot Soccer System Based on Global Vision System,TP242.6
  7. Research on Multi-Robot Cooperative Pursuit Problem,TP242
  8. Reaserch on Dynamics for Fast Walking of Biped Robot,TP242.6
  9. The Research on a New Modular Self-Reconfigurable Robot,TP242.2
  10. Mechanism Design and Kinematic Simulation of a Celiac Invasive Surgical Robot,TP242.3
  11. Key Control Technique of Wafer Handling Robot,TP242.2
  12. Training Program and Rehabilitative Effect Appraisal of Lower Limbs Rehabilitative Robot,R49
  13. Humaniod Robot Development and Research,TP242
  14. Design and Simulation on MDOF Robotic Prosthesis’s Shoulder,TP242
  15. Study of the Mechanical Strcture of a De-Icing Robot for 220kV Transmission Lines,TP242
  16. Algorithm Research for Robot Trajectory Tracking Control,TP242
  17. Study on Robot Welding Technique for Double-hull Ship Block,TP242
  18. Research on the Complete Coverage Path Planning Algorithm of Mobile Robot,TP242
  19. Study of the Structure of Wall-Inspecting Robot on Chemical Container in Service,TP242
  20. Study on the Technique of Information Fusionapplied to Enbedded Driver Fatigue Detection,TP368.12
  21. Research on Navigation System of Agricultural Wheeled-Mobile Robot Using Multi-Information Fusion,TP242

CLC: > Industrial Technology > Automation technology,computer technology > Automation technology and equipment > Robotics > Robot > Industrial robots
© 2012 www.DissertationTopic.Net  Mobile